The Political Consequences of the Great Recession in Southern Europe Crisis and Representation in Spain

Guillem Vidal

Thesis submitted for assessment with a view to obtaining the degree of Doctor of Political and Social Sciences of the European University Institute

Florence, 13 June 2019 ii European University Institute

Department of Political and Social Sciences

The Political Consequences of the Great Recession in Southern Europe Crisis and Representation in Spain

Guillem Vidal

Thesis submitted for assessment with a view to obtaining the degree of Doctor of Political and Social Sciences of the European University Institute

Examining Board Prof. Hanspeter Kriesi, European University Institute (Supervisor) Prof. Elias Dinas, European University Institute Prof. Eva Anduiza, Universitat Autònoma de Barcelona Prof. Kenneth M. Roberts, Duke University

© Guillem Vidal, 2019 No part of this thesis may be copied, reproduced or transmitted without prior permission of the author

Researcher declaration to accompany the submission of written work Department of Political and Social Sciences - Doctoral Programme

I Guillem Vidal certify that I am the author of the work The Political Consequences of the Great Recession in Southern Europe: Crisis and Representation in Spain I have presented for examination for the Ph.D. at the European University Institute. I also certify that this is solely my own original work, other than where I have clearly indicated, in this declaration and in the thesis, that it is the work of others. I warrant that I have obtained all the permissions required for using any material from other copyrighted publications. I certify that this work complies with the Code of Ethics in Academic Research issued by the European University Institute (IUE 332/2/10 (CA 297). The copyright of this work rests with its author. Quotation from it is permitted, provided that full acknowledgement is made. This work may not be reproduced without my prior written consent. This authorisation does not, to the best of my knowledge, infringe the rights of any third party. I declare that this work consists of 63.058 words.

Statement of inclusion of previous work: I confirm that chapter 2 was jointly co-authored with Björn Bremer and I contributed 50% of the work and draws upon an earlier article published in Evdoxios Doxiadis and Aimee Placas (eds.), Living Under Austerity: Greek Society in Crisis, vol. New York: Berghahn Books, 113– 140.

I confirm that chapter 3 was jointly co-authored with Dr. Swen Hutter and Dr. Hanspeter Kriesi and I contributed 33% of the work and draws upon an earlier article published in Party Politics (2018), 24:1, 10–22.

I confirm that chapter 4 draws upon an earlier article I published in West European Politics (2018), 41:2, 261–286.

I confirm that chapter 5 was jointly co-authored with Irene Sánchez-Vítores and I contributed 60% of the work.

Statement of language correction: This thesis has been corrected for linguistic and stylistic errors. I certify that I have checked and approved all language corrections, and that these have not affected the content of this work.

Signature and date:

v vi To my family and friends, without whose never-ending entertainment this dissertation would have been finished in half the time.

vii

viii Table of Contents

Abstract ...... xi

Acknowledgements ...... xiii

List of tables ...... xvii

List of figures ...... xix

Chapter 1: Crisis Dynamics and Political Transformations: An Introduction ...... 1

Chapter 2: From Boom to Bust: A Comparative Analysis of Greece and Spain under Austerity ...... 17

Chapter 3: Old versus new politics: The political spaces in Southern Europe in times of crisis ...... 47

Chapter 4: Out with the Old: Restructuring Spanish Politics ...... 75

Chapter 5: Challenging Business as Usual? The Rise of New Parties in Spain in Times of Crises ...... 103

Chapter 6: Conflict Activation in Times of Crisis: The Case of Catalonia ...... 131

Chapter 7: Conclusion: Towards a new structure of conflict in Southern Europe? ...... 163

Bibliography ...... 169

Appendices ...... 185

ix x Abstract

The Great Recession constituted a breaking point in several aspects of the cultural, economic and political life of southern European countries (i.e. Greece, Italy, Portugal and Spain). This dissertation aims to shed light on the political consequences of the economic crisis in this region —with a specific focus on Spain as a paradigmatic case— by analysing different aspects of the political transformations that took place during the period of crisis. The underlying argument is that, albeit some relevant differences, the four countries experienced a common pattern: the incapacity of national politics to offer differentiated recipes to the deteriorating economic situation triggered a widespread crisis of representation that introduced new issues in the political agenda and drove the political transformations in these countries. The combination of a political and economic crisis at the national and European levels opened new political spaces that new parties capitalised by appealing to the need for democratic renewal and opposition to austerity politics. Furthermore, as illustrated by the Spanish case, and in particular the Catalan experience, the political crisis had far-reaching consequences beyond economic grievances, leading to the activation of different types of conflicts. Overall, the findings suggest that the transformations in the structure of political conflict in southern Europe in the aftermath of the Great Recession are not the by-product of a growing cultural divide —as is the case in several other continental and north-European countries—, but instead respond to the loss of credibility in the political system. Methodologically, the dissertation relies on an original dataset of media content as well as on several sources of survey data to test the empirical validity of the claims.

xi xii Acknowledgements

The work that follows, as any intellectual exercise, is inextricable from one’s own life experience. If the topic of this dissertation is about the political consequences of the Great

Recession in southern Europe with a focus on Spain, it is because, as a young student at the time the Great Recession broke out, I witnessed first-hand the political and social reaction that the crisis triggered in my immediate environment and became profoundly captivated by it. I had to know more about what this social and political earthquake meant. Captivation turned into curiosity, and curiosity into questions. If I have had the privilege to dedicate the last five years at the European University Institute —with a stay in the Carlos III University in

Madrid— to reflect and analyse some of these questions in an exceptional cultural and intellectual environment, it is thanks to the many people who have supported me and influenced my thinking in a myriad of ways. To all of them I owe any of the good that the reader might find in this dissertation. The words that follow, even if they inevitable fall short to express all of my gratitude, are dedicated to them.

Several of the underlying ideas behind this work are strongly influenced by the collective endeavour of the POLCON (Political Conflict in the Shadow of the Great Recession) project.

Each of the members of the team have contributed greatly to my intellectual development over the years. The long discussions, the conferences, the endless hours spent on planning, writing, analysing and coding, even the disagreements, have shaped each of us, without a doubt, into who we are today, both personally and professionally. The project, which would have not been possible without Maureen – our unconditionally supportive secretary and friend

– has been an unforgettable challenge. Certainly not always easy, but enormously rewarding on many different levels.

I would like to thank my supervisor, Hanspeter Kriesi, for his intellectual mentorship over the years. The influence that his thinking and method of working has had on me is impossible to

xiii overstate. For that, and for allowing me the freedom to explore my own path while providing valuable feedback and advice, I am eternally grateful.

I would also like to thank my unofficial second supervisor and friend, Swen Hutter, for becoming an indispensable source of encouragement during the difficult times when the purpose and direction of the research seemed lost, and generously dedicating his time to helping me structure many of the ideas behind this work. Without his personal and professional support, I cannot say for sure that I would have brought this dissertation to a conclusion.

If I managed to pursue a Ph.D. it is without a doubt thanks to my master’s thesis supervisor and friend Diego Muro. Had it not been for his encouragement and support at the early stages of my career, I would not be writing these words. Finding someone who believes in your capacity, pushes you to develop your abilities, and provides you with an opportunity to develop a career, is nothing short of a treasure.

I would, at this point, also like to thank the other co-authors of the pieces included in the dissertation, Björn Bremer and Irene Sánchez-Vítores. Collaborative efforts are an exceptional way to learn about oneself and broaden perspectives. I am very glad to have had the opportunity to work with these two excellent researchers and thank them for their tireless effort. I would also like to mention the many colleagues and friends from Florence and abroad

– some of them who are also co-authors in different projects – who have kindly taken the time to read and provide feedback of several parts of this dissertation: Pablo Simon, Siljia

Hauserman, Reto Burgisser, Giorgio Malet, Martijn Vlaskampf, Berta Barbet, and Jorge

Galindo, amongst others. Without them the quality of this work would not be nearly the same.

The last words I want to dedicate to my friends and loved ones, some of them whom I met during this extraordinary period and whom I hope to keep close to for the rest of my days.

xiv To Tello, my brother from another mother and father. To Simone, my flatmate and comrade in the battle for surviving the writing of a Ph.D., and the person guilty for my strange Milan slang in Italian. To Giorgio, for Napoli, and Wiebke, for the madness, and all of the unforgettable moments. To Matthieu, Paula and Koen, for the po’ più di vinos, and to all of the musicians who have made the jam sessions and other concerts possible in Fiasco these past years (rock on Patrice, Diego, Oscar, Maria, Pagano, Edgars, and long live Sugar Plum

Fairies!). To all of ‘Divertimento Serale’ & ‘Turbocane Barbatua’1. To Scipione Ammirato.

To Quelo. To Selva de Mar. To Petersmann and cinema. To Antoni (aka cremis), Pablo (aka

Notas), Carlos (NP), Luís, Arcarons, and Nacho, for all of the invaluable laughs. To Giorgia, for Rome, hopefully Moscow, and all the good times. To Albert, for making me a better person. To Reto, for Amrì-Piotta, Berlino, and all of the good times still to come. To Julian

(aka Pyro), for all the burning. To the friends that live far yet they are always close: Maayan,

Max, Johannes, Jerome, Lucas, Martí, Bruno, Laia (and Abril). To everyone else I might be forgetting as I type these words late in the night.

To my brothers, Felipe (plus naranja esplotá) and Gerard (aka The Bird Yellow), for all you’ve taught me. To all of my grandparents, for their kindness, inspiration, poetry, and love.

To my uncles and extended family, for should every human being be so lucky to have such affective and supportive network.

To my father, Ramon, for having taught me a lot of the little I know, and for the unconditional love and support demonstrated throughout the years.

To my mother, Isabel, to whom I owe my awakening to the beauty of culture and intellectual curiosity.

1 It should be noted that the name of this now famous football club comes from a misunderstanding of the Spanish words ‘Tobogán’ and ‘Barbacoa’ as mistakenly understood by Simone Cremaschi in the late hours of the Florentine night.

xv

xvi List of Tables

Table 1.1 Outline of the Dissertation ...... 2

Table 2.1 Gini Coefficient in Germany, Greece, and Spain, 2005–16 ...... 32

Table 2.2 Spain General Elections Results, 2008–16 ...... 37

Table 2.3 Greece General Elections Results, 2009–15 ...... 38

Table 3.1 (Expected) Conflicts in Southern Europe ...... 51

Table 3.2 Issue salience by major issue categories in Southern Europe ...... 61

Table 4.1 Election results and party system features: percentages and index results ...... 80

Table 5.1 Parties and Number of Observations in the CIS survey, 2016 ...... 114

Table 5.2 Marginal Effects of Multinomial Logistic Regressions on Vote ...... 120

Table 6.1 Types of displacement strategies by issues in party competition ...... 129

Table 6.2 Catalan Parliament Election Results (percentage over valid votes), 1984-2017 ... 135

Table 6.3 Marginal effects of leavers versus loyalists and newcomers vs. opposers of CiU, 2010-2012 ...... 147

xvii xviii List of Figures

Figure 2.1 Financial Account as a pct. of GDP of Germany, Greece, and Spain, 2000–16 .... 21

Figure 2.2 Unemployment Rate in Germany, Greece, and Spain, 2000–16 ...... 31

Figure 2.3 Satisfaction with the European Union and National Democracy, Greece and Spain, 2000–16 ...... 34

Figure 2.4 Trust in National Institutions, Greece and Spain, 2000–16 ...... 35

Figure 2.5 Trust in European Institutions, Greece and Spain, 2000–16 ...... 35

Figure 3.1 Issue salience and polarization by specific issue categories ...... 63

Figure 3.2 Political spaces in Southern Europe, 2011 to 2015 ...... 66-67

Figure 4.1 Salience of issues by election in Spain ...... 88

Figure 4.2 Position of parties by election on main issues in Spain ...... 91

Figure 4.3 Political Space in the Spanish campaigns of 2004, 2008, 2011 and 2015 ...... 93

Figure 5.1 The Emergence of Podemos and Ciudadanos in the Polls ...... 101

Figure 5.2 Political and Economic Crisis Indicators in Spain, 2000-2016 ...... 107

Figure 5.3 Average position of voters of main parties in Spain according to various ideological dimensions ...... 117

Figure 5.4 Average Marginal Effects of Interaction between Political Crisis Indicator and Ideology on Probability to Vote for PP, PSOE, Podemos, and Ciudadanos ...... 121

Figure 5.5 Average Marginal Effects of Interaction between Political Crisis Indicator and Age on Probability to Vote for PP, PSOE, Podemos, and Ciudadanos ...... 122

Figure 6.1 Evolution of Catalans opinion with regards to the relation between Spain and Catalonia ...... 132

Figure 6.2 Share of pro-independence by party voted in last elections, 2006-2014 ...... 144

Figure 6.3 Predicted probabilities of vote for independence by preference for the territorial organisation by political grievances ...... 149

Figure 6.4 Predicted Probabilities of voting for a pro-secessionist party (CiU, ERC and CUP, or coalition JxSi/JxCat) by national identity in different elections ...... 151

Figure 6.5 Predicted Probabilities of abstaining by preferences over territorial organisation in different elections ...... 152

xix xx Chapter 1. Crises Dynamics and Political Transformations: An Introduction

Introduction

In the Great Transformation, Karl Polanyi suggested that there are critical periods when ‘time expands’. A detailed analysis of these periods is valuable to understand the dynamics of social and political transformations. By all accounts, time ‘expanded’ in southern Europe in the shadow of the Great Recession. The multiple crises that unfolded after the fall of Lehman

Brothers in 2008 constituted a breaking point in several aspects of the cultural, economic and political life of these countries (i.e. Greece, Italy, Portugal and Spain). The aim of this article- based dissertation is to shed light on different aspects of the political transformations that unfolded in this region in the aftermath of the crisis.

The underlying argument is that, albeit some relevant differences, the four countries experienced a common pattern: the incapacity of national politics to offer differentiated recipes to the deteriorating economic situation triggered a widespread crisis of representation that introduced new issues in the political agenda and drove the political transformations in these countries. The combination of a political and economic crisis at the national and

European levels opened new political spaces that new parties capitalised by appealing to the need for democratic renewal and opposition to austerity politics. Furthermore, as illustrated by the Spanish case, and in particular the Catalan experience, the political crisis had far- reaching consequences beyond economic grievances, leading to the activation of different types of conflicts. The findings suggest that the transformations in the structure of political conflict in southern Europe in the aftermath of the Great Recession are not the by-product of a growing cultural divide —as is the case in several other continental and north-European

- 1 - countries (Kriesi et al. 2008, 2012)—, but instead respond to the loss of credibility in the political system, which interacted with pre-existing lines of domestic conflict to produce new political configurations.

In this introductory chapter I briefly introduce the broad theoretical underpinnings that serve as a basis for the more detailed theoretical sections in each of the chapters of the dissertation.

Before that, in the following section, I present the outline and briefly summarise the contributions of each of the individual chapters.

Thesis Outline

The dissertation is composed of two comparative articles and three case study chapters focusing on Spain, which is taken as a paradigmatic case of transformation where two new parties emerged in the years following the Great Recession. While some of the chapters speak to each other, it should be noted that these are individual pieces of work that tackle different specific aspects of the broader political transformations in south European countries. Table 1 shows the overall structure of the dissertation.

Table 1.1 Outline of the Dissertation

Chapter 1. Crises Dynamics and Political Transformations: An Introduction

Chapter 2. From Boom to Bust: A Comparative Analysis of Greece and Spain under Austerity with Björn Bremer Chapter 3. Old versus new politics:

Comparative The political spaces in Southern Europe in times of crisis with S. Hutter and H. Kriesi

Chapter 4. Out with the Old: Chapter 5. Challenging Business as Usual? Restructuring Spanish Politics The Rise of New Parties in Spain in Times of with I. Sánchez-Vítores Crises Supply-side Demand-side Chapter 6. Conflict Activation: Case Study: Spain The Case of Catalonia

Chapter 7. Conclusion

- 2 - The first article, Chapter 21, reviews the economic and political origins and consequences of the crisis that started in 2008 in comparative perspective in Greece and Spain. Since the remainder of the dissertation focuses on the political consequences of the crisis, this chapter fits well into an introductory role insofar it covers the political economy of the Eurozone in the context of the crisis and shows the great economic dislocations suffered by these countries. It is contended that prior economic and political developments are fundamental in understanding the economic crisis that unfolded in 2008. In particular, irresponsible governments in Spain and Greece failed to regulate the pre-crisis boom effectively. Driven by incentives that derived from the ill-conceived institutional structure of the EU, they created a socio-economic environment, in which the economic crisis could only have the disruptive political consequences that we have witnessed.

Chapter 32 then engages with the changes in the political debate and the position of the actors in the elections after the crisis in comparative perspective. It is argued that the configuration of conflict in these countries was shaped by an economic crisis that coexisted with a political crisis that had both European and domestic dimensions. These changes were led by social movements and political parties that forcefully combined opposition to austerity and to ‘old politics’. While this pattern emerged everywhere, there are distinct county differences. To advance the understanding of those differences, the rest of the dissertation — chapters 4, 5 and 6— focus on Spain as a paradigmatic case of transformation and aim to disentangle the mechanisms behind these political changes.

1 Published as: Bremer, Björn, and Guillem Vidal (2018). ‘From Boom to Bust: A Comparative Analysis of Greece and Spain under Austerity’, in Evdoxios Doxiadis and Aimee Placas (eds.), Living Under Austerity: Greek Society in Crisis, vol. New York: Berghahn Books, 113–140. 2 Published as: Hutter, Swen, Hanspeter Kriesi, and Guillem Vidal (2018). ‘Old versus new politics: The political spaces in Southern Europe in times of crises’, Party Politics, 24:1, 10–22.

- 3 - More concretely, Chapter 43 relies on an original dataset of media content to analyse party’s positions and overall transformations in the Spanish political landscape, covering two elections before the beginning of the crisis (2004, 2008) and two elections afterward (2011,

2015). The results suggest that, while the elections before the crisis were characterised by two poles identifiable on the left-right scale and cultural and territorial issues dominated during this period, the crisis brought about a multipolar structure of conflict in which economic and political issues dominated the public debate and structured party competition.

In this changing political landscape, Chapter 54 explores the collapse of the Spanish party system and the reasons behind the emergence of two new parties: Podemos and

Ciudadanos. Using survey data to map the ideological space and model voting behaviour, it is shown that economic voting is only party of the story. The chapter contends that the transformations in the Spanish party system are best understood through the prism of the crisis of representation that unfolded alongside a severe economic crisis: it is the dissatisfaction with the political system which drives the vote for both new parties. The results also show that a unidimensional structure and a generational divide cut across these critical attitudes: the young and the politically dissatisfied are more likely to vote for new parties, each on different sides of the political spectrum.

While there are common patterns across south European countries in the types of conflict that emerged in the shadow of the crisis, it is crucial to keep in mind that the resulting configurations of conflict are contingent upon pre-existing structures of conflict. Chapter 65

3 Published as: Vidal, G. & Sánchez-Vítores, Irene. 2018 "Out with the Old: Restructuring Spanish Politics" in Restructuring European Party Politics in Times of Crisis. eds. Swen Hutter and Hanspeter Kriesi. Cambridge: Cambridge University Press. 4 Published as: Vidal, Guillem (2018). ‘Challenging business as usual? The rise of new parties in Spain in times of crisis’, West European Politics, 41:2, 261–286. 5 Unpublished manuscript.

- 4 - zooms into the Catalan secessionist challenge to explore the dynamics of conflict activation in the aftermath of the crisis related to the territorial dimension of Spanish politics. The article traces the Catalan case to show that, following the 2010 elections, a process of ethnic polarisation led to an ethnic alignment of the vote in subsequent elections (2012, 2015, 2017).

This process saw changing positions on fundamental issues by mainstream parties which the article analyses, suggesting that conflict-displacement strategies due to adverse economic circumstances can account for such dynamics.

Finally, chapter 7 presents some concluding remarks, speculates about the future of south European politics, and raises questions for future research.

Political Conflict and Party System Change: A Theoretical Framework

There are two fundamentally distinct yet simultaneous processes that must be accounted for in understanding party system change in Europe over the last decades. The first refers to the long-term process of cleavage (trans)formation that has reshaped north-western European

(NWE) party systems with the emergence of a new cultural (or new politics) dimension since the 1970s. Political sociologists have devoted significant effort to understanding the emergence and consolidation of this new divide and its political translation (Kriesi et al.

2008, 2012). Essentially, the emergence of the Populist Radical Right (PRR) and the New

Left are often explained on the basis of this new cultural cleavage deriving from the social transformations brought about by globalisation. The second one refers to the growing amorphousness of the electorate and the decline of enduring social constituencies that has led to increasing disengagement from politics, representing the downfall of traditional party politics (Mair 2013).

While the two processes must be understood as complementary, I argue that each of them explains different types of changes in the party systems across Europe’s regions. On the one hand, the PRR and the New Left have progressively occupied a ‘niche’ space in a new

- 5 - cultural cleavage opened in NWE due to the social transformations deriving from globalisation and technological change. On the other hand, the populist left parties in

Southern Europe (SE) are the consequence of the heightening of the tensions that stem from the constraints on national sovereignty and a perceived decline of representativeness resulting from the economic crisis, which mirrors the second process. This distinction is fundamental in understanding the regional variation in the transformations in European political spaces. The next part reviews the evolution of cleavage structures in Europe and stresses the shortcomings for explaining the current political alignments in certain European regions. The following part deals with the role of parties in government, introduces the literature on the crisis of representation and develops the argument that economic crisis can act as a heightener of the tensions between responsibility and responsiveness, leading to the collapse of party systems.

Cleavages and Realignment

No literature on cleavage theory in the last decades can avoid dealing with the seminal work of Lipset and Rokkan (1967b). In a historical account of Western Europe, going back as far as the Middle Ages, their work inquired about the historical conditions and development of cleavages and the transformation of sociocultural conflicts into opposition between parties.

Borrowing the words of Satori (1969) —who two years later would already describe their work as a “landmark” that wold “rebalance the discipline” of political sociology—, Lipset and

Rokkan dealt with the question of “how are conflicts and cleavages translated into a party system?” (Sartori 1969: 207). An in-depth review of this classic work is far beyond the scope of the present document (see Bornschier 2010). However, there are a few important aspects to highlight for the purpose that concerns us here, which is an overview of the current development of contemporary cleavages in western Europe.

Perhaps the most well-known and cited idea of Lipset and Rokkan’s (1967b) work is that of the freezing hypothesis. In short, the authors introduced the idea that party systems in

- 6 - the 1960s displayed a remarkable similarity to the party systems already set in place in the

1920. Several works have addressed this statement and its empirical validity over the years

(see Kriesi 1998; Bornschier 2010). Perhaps one of the most influential works in that regard is that of Mair and Bartolini (1990) and Mair (1997a) who empirically address the patterns of electoral change and party competition until the 1990s and conclude that there is a strong persistence of cleavages. In Mair’s own conclusion of Chapter 4 titled the Myths of Electoral

Change, he argues: “following Rokkan, the party alternatives of the 1960s were older than the majority of their national electorates. Thirty years on, these self-same parties still continue to dominate mass politics in western Europe. Nowadays, in short, they are even older still”

(Mair 1997a: 90).

The two main critical processes that reshaped and later stabilized party systems in the

1920, according to Rokkan’s (1999) analysis, were the National revolution and the Industrial

Revolution. These two processes created four types of cleavages, two related to each of the revolutions. The National revolution, on the one hand, brought about conflicts over values and cultural identities, reflected in the religious/secular and centre/periphery cleavages

(Rokkan 1999: 284). The Industrial revolution, on the other hand, brought about “conflict between the landed interests and the rising class of industrial entrepreneurs; the conflict between owners and employers on the one side and tenants, labourers, and workers on the other” (Rokkan 1999: 284). In short, the capital/labour and urban/rural cleavages.

Have these ‘classical’ cleavages maintained the same strength in predicting vote and structuring party systems since then? Several works have sought to address this question pointing to an attenuation of both the religious and class cleavages —or at least its transformation— which were considered the two most important determinants of party choice

(Lijphart 1979). Given the extensive social transformations that have occurred since the

1960s, it is no wonder that the class cleavage has transformed. There are a number of factors

- 7 - that have contributed to these social transformations that have rendered a much more diversified and complex social structure. The process of deindustrialisation, feminization and tersiarisation of the labour market, together with the expansion of the welfare state and increasing levels of education are all factors that play a role in the redefinition of class, and therefore, its linkage to class voting (see Oesch 2006b; Evans and Graaf 2013). To put it bluntly, social class can no longer be approximated dichotomously as in the classical Marxist terms of the bourgeois and the proletariat, since the divisions no longer represent the amorphousness of the labour market (Evans and Tilley 2017; Wright 2015). However, this does not necessarily imply that class voting is no longer relevant. The shrinking sheer size of the working class simultaneously saw the emergence of a new middle class (Kriesi 1998).

Thus, class voting might have not necessarily declined, but rather social classes transformed, and a redefinition is needed to understand the new version of the class cleavage voting (Oesch

2006b)6. To reflect that change, a new strain of literature has followed a more ‘Weberian’ approach to the definition of social class, i.e. focused on labour market relations (see

Connelly et al. 2016).7

Similarly to the traditional class cleavage, religious affiliation has been argued to have become “a less important guide for political choices” due to a process of “secularization and individualisation” (Evans and Graaf 2013: 5). Although religion in itself seems to have taken a back seat in the structure of political conflict, most literature still refers to the ‘cultural’

6 In that sense, Inglehart and Flanagan’s (1987: 1297) argument that “the rise of postmaterialistic issues […] tends to neutralize political polarization based on social class” is somewhat misconceived. For Inglehart, a new ‘value cleavage’ had replaced traditional cleavage structures since the end of World War II (Inglehart 1997, 1990, 2015). Yet, this new ‘value cleavage’ might have not necessarily replaced the class cleavage but simply overlapped with a transforming class structure. 7 Several works have focused on the empirical relation between voting behaviour and this new type of social class in what could be qualified as a renaissance of class voting (see Oesch 2012; Oesch and Rennwald 2017; Rennwald and Evans 2014).

- 8 - cleavage as the fundamental driver of the contemporary structure of conflict, which is not independent from the historical role of religion. As Rovny and Polk (2014) suggest, the socio- cultural dimension is in fact intrinsically linked to the historical legacies of the religious conflict in Europe. Yet, the content of the cultural dimension is no longer solely composed of the religious issue but with issues that derive from the most important process that reshaped party conflict in the last decades: globalisation.

While the formation and stabilisation of cleavages described by Rokkan had a lot to do with the nation-building process and the nationalisation of politics, much of the contemporary literature on cleavages is concerned with the impact of globalisation in opening up a new cultural dimension of conflict8. Several authors have theoretically argued and showed empirically the emergence of this new cultural cleavage (e.g. Inglehart 1997;

Kitschelt 1988). Kriesi et al. (2006, 2008, 2012) identify two waves of transformation that reshaped North-European party systems – a first wave led by the new left in the 1970s and early 1980s, and a second wave driven by the PRR since the 1990s. The new left advocated for the protection of the environment, the free choice of lifestyles and other universalistic values (e.g., della Porta and Rucht 1991; Kitschelt 1988). The PRR, by contrast, focused on immigration and European integration (see Hutter et al. 2016; Hooghe et al. 2002) as threats to the homogenous nation-state to be confronted. These two new families of parties would occupy opposing spaces in a new cultural dimension (integration-demarcation) that Kriesi et al. argue that coexists with a ‘classical’ economic dimension concerning the traditional socialist (more redistribution) vs capitalist (more market) dimension. As already pointed out by Kriesi (2010: 683), this cultural cleavage has received several names in the literature.

8 Paradoxically, while the process of ‘denationalisation’ is argued to be the key driver of this new ‘critical juncture’, “the political reactions to […] globalisation are bound to manifest themselves above all at the national level.” (Kriesi et al. 2008: 3). Thus, while globalisation may very well be the key driver of change in contemporary European societies, the conflict itself is integrated into national politics.

- 9 - Hooghe et al. (2002), Marks et al. (2006) called it the ‘new political dimension’, referring to the two pols as GAL (Green, Alternative, Libertarian) and TAN (Traditional, Authoritarian,

Nationalist). Moreover, Bornschier (2010) refers to it as the ‘libertarian-universalistic / traditionalist-communitarian cleavage’. Beramendi et al. (2015a) call it the ‘universalism vs particularism’ division.

This process is, however, by no means homogeneous. Although this general framework seems to accurately depict the transformations across several western European countries, it far from captures the recent transformations across all European countries. In fact, most of the above-mentioned literature has focused exclusively on NWE countries. The transformations in post-communist countries in CEE (Kitschelt 1995; Marks et al. 2006; see

Rohrschneider and Whitefield 2012) and SE (Hutter et al. 2018) have undergone different transformations altogether. In both of these regions, equivalent transformations (i.e. the two waves) did take place in the same way, and as such, did not result in a stable two-dimensional structure of conflict. Apart from the belated democratization of CEE and SE countries (with the exception of Italy), their party systems were also less institutionalised, arguably producing a configuration of parties less rooted in social divides. With the Great Recession (GR) that started in 2008, these countries have undergone substantial political and economic transformations. Comparative work on the transformations of these three regions argues that, while in CEE cultural conflicts —especially related to nationalism— have continued to dominate political conflict in rather unstable party systems, changes in SE countries —albeit important regional differences— have been dominated by a new dimension of conflict related to political renewal and the regeneration of politics (see Chapter 3).

The transformations of the party systems that took place in the aftermath of the GR in

2008, suggest that long-term cleavage transformations cannot exclusively account for party system change. In the next section I dwell on the concept of a crisis of representation as

- 10 - another approach to understand party system transformations under the impact of an economic shock that relates to the dynamics of parties in governments.

Crisis of Representation and Party System Collapse

In linking cleavage theory with theories of political realignment, Martin (2000; in Bornschier

2010: 4) distinguishes between three levels of analysis. The first concerns the “long-term evolution of social structure, [in which critical junctures] create structural potentials that political actors can mobilize”. The second one —at the intermediate level—, is concerned with “the adaptation of the existing structure of conflicts to new potentials, […] and the way for the establishment of new links between social groups and political parties. Finally, the last level focuses on “everyday politics”, such as corruption scandals, and which “they rarely affect the two higher levels of political development.” (Bornschier 2010: 4). So far, the driver of party system restructuration discussed above has been that of linking the first and the second levels of analysis, that is, how long-term cleavages change the structural potentials of conflict that political actors can mobilise. However, this approach can hardly grasp the dramatic changes that shook the SE party systems in the aftermath of the GR, where a political crisis unfolded alongside the economic crisis. Instead, these transformations might be related to another long-term process —and its short-term impact— that concerns the growing difficulty of parties in government to effectively represent the demands of the electorate during an economic crisis (Mair 2013, 2009).

Let us begin with the definition and identification of a political crisis or crisis of representation, and the link to economic crises. Democratic representation consists of a principal-agent relationship between the voters (the principals) and the politicians (the agents) that establishes channels for interests to be represented institutionally. Elections are thus the mechanism whereby voters put trust in politicians hoping they will defend these interests, which might take different forms, such as delivering material advantages (clientelism) or the

- 11 - pursuit of policy programmes (programmatic) (Kitschelt 2000; Kitschelt and Wilkinson

2007). It is precisely through the failure of these linkages between the electors and parties that a crisis of representation occurs and a window of opportunity for new parties opens (Morgan

2011). From a theoretical standpoint, the failure of linkages between voters and politicians will destabilise party competition. But how are we to recognise a crisis of representation?

Mainwaring (2006) distinguishes two dimensions that constitute a political crisis or crisis of representation: an attitudinal dimension, involving citizens’ perceptions; and a behavioural dimension, which encompasses ‘actions by citizens rejecting existing mechanisms of democratic representation’ (Mainwaring 2006). The attitudinal dimension, also referred to as latent political potentials (Kriesi 2015: 3), constitutes the degree to which citizens are dissatisfied with the political system and (do not) feel represented by democratic institutions. It is therefore an indicator of the extent to which there is fertile ground for a political crisis to emerge. The latter only erupts when it is manifested in the behavioural dimension ‒ that is, when there is an overt manifestation (Kriesi 2015). The behavioural dimension is concerned with the manifestations in both the formal (electoral volatility, changes in the party system, emergence of new parties) and informal (protests) political arenas.

But what is it that produces a crisis of representation? As first pointed out by Krastev

(2002) for the Balkans, and later applied to Southern Europe by Bosco and Verney (2012) — though it would also apply to several Latin-American cases—, it is the situation of a

‘democracy without choices’ which produces frustration with the democratic process and aggravates the crisis of political representation. This situation is the outcome of what is best described by the growing tension that governments face between acting ‘responsible’ versus

‘responsive’ (Mair 2009), or in other words, in dealing with the gap between democracy and efficiency (Scharpf 1999). Under the external constraints of supranational institutions,

- 12 - governments were pressured into adopting a set of fiscal consolidation policies and structural reforms with the objective of reducing debt and regaining access to international credit markets. Incumbents of all sides of the ideological spectrum introduced these unpopular measures to different degrees claiming to act ‘responsibly’ at the cost of being, especially for the left, ‘responsive’ to both their constituencies and ideological principles.

Focusing on the Latin-American experience, Roberts (2013) shows that the restructuring of party systems was largely contingent upon the implementation of neoliberal market reforms during the regions’ economic shock in the 80s and 90s. The reforms introduced, as well as the constraints placed on governments under the so-called Washington

Consensus, hold suggestive resemblances to the implementation of austerity reforms in

Southern Europe (Blyth 2013). Roberts distinguishes between three types of critical junctures that led to different outcomes: 1) those that aligned party systems programmatically; 2) those that dealigned them; and 3) those that had a neutral effect (Roberts 2013: 1436). In the cases where the conservative governments implemented reforms with the opposition of the centre- left, programmatic alignment occurred. This is referred to as a situation of contested liberalism. On the contrary, when major political contenders participated in the process of market liberalization, that is, when centre-left governments either supported or directly introduced reforms in line with the neoliberal orthodoxy, then programmatic dealignment occurred, leading to a situation of neoliberal convergence. The dealignment critical junctures, thus, would be expected to produce ‘popular resistance […eventually] channelled into institutionalized patterns of representation’ (Roberts 2013: 1438). In other words, in those countries where centre-left parties partook in the implementation of neoliberal reforms, the stability of the party systems is expected to be jeopardized, opening space for protests and new populist parties on the left. In those countries where the structure of competition remained aligned, it is expected that the conflict over market reforms intensified thus

- 13 - strengthening the position of both main competitors. In a similar fashion, Lupu (2014, 2016) fundamentally argues that the situation of neoliberal convergence produces a dilution of the brand of the party. It is therefore not the incumbent performance but the blurring of the competition of the parties from the left and the right in a context of international economic constraints that is responsible for the party breakdown in the region.

Weather we focus on the EU’s constraints on labour and economic policies in Europe or the case of Latin-America’s during the Washington Consensus, the difficulty for parties in government to reconcile the preferences of the electorate with the demands of international actors (i.e. the so-called the gap between responsibility and responsiveness) is heightened during economic crises. This situation arguably destabilises party competition and opens structural potentials for the emergence of new parties. There is an important link to be made here between the theory of responsibility and responsiveness and theories of international political economy, which link it to the process of globalisation. In chapter 9 of The

Globalisation Paradox, Dani Rodrik (2013) presents a framework to understand the tensions that stem from a globalising economy which he calls The Political Trilemma of the World

Economy (2013: 201). The global economy, according to Rodrick, can be simplified with a model dictating a set of impossibilities represented by three poles: 1) deep economic integration (or Hyperglobalisation), 2) nation state and 3) democratic politics. Any of these two combinations can go together excluding the third option. Thus, global economic integration can be integrated with national sovereignty at the cost of democracy, while combining democratic politics and national sovereignty would result in giving up global financial integration.

Peter Mair’s work on the growing gap between responsibility and responsiveness complements well with this theory to explain why a crisis of representation unfolds.

Essentially, in times of economic turmoil, the incompatibilities between satisfying the

- 14 - requirements to remain integrated in the global world economy and the demands that would satisfy national interests become more accentuated, forcing political parties to adopt one or the other. It is no surprise that this situation is particularly accentuated in southern Europe where the crisis hit hardest. In these countries, the reforms by the EU pursued an economic integration agenda, pushing governments to introduce unpopular reforms and cuts in the welfare systems. These neoliberal reforms targeted at efficient economic integration, thus, would pursue the economic integration and nation-state duet at the cost of democratic politics in Rodrick’s trilemma, and in words of Peter Mair, the parties supporting this strategy would be acting responsible towards a set of external actors that would come at the cost of responsiveness, or democratic politics. Matthijns (2017) introduces this very framework applied to the EMU’s context to understand the political changes in Southern Europe.

The new parties that benefit from the widening gap between satisfying the interests of the external actors and national interests claim that they will be responsive to the “people” as opposed to “responsible” to the external, undemocratic, actors. This is, in fact, the key driver of a second fundamental process of the transformation of European politics: populism.

Although often attached to specific cultural issues or values such as anti-immigration, national attachment, or opposition to European integration, populism is independent from specific issues and is merely concerned with the representativeness gap that emerges from party’s incapacity to alleviate the growing constraints placed on governing actors. While this process is often framed as a new phenomenon, Dahl’s (1956) Preface to Democratic Theory

(see also Dahl 2000) already introduced two models of democracy that reside at the core of these transformations, namely the populist and Madisonian (or liberal) types or visions of democracy. In a context where voters do no longer feel attached to the available political choices, the populist version of democracy is bound to be picked up by a new challenger.

- 15 - To summarize, while the transformations in both the NWE party system and SE are ultimately linked to globalization and its constraints, the mechanisms and the reasons for these transformations are fundamentally different. In essence, PRR parties emerging in the

80s in NWE respond to a long-term process of transformation of cleavage structures with the emergence of a new value divide (and its opposition). Globalisation would have created

‘winners’ and ‘losers’, eventually taking the shape of a cultural divide in the political supply.

These parties would have progressively ‘creeped’ into the party systems by exploiting the new divide. The story for the populist left parties and the link to globalisation in SE is, however, different9. Instead, these parties are the product of the constraints placed on party competition during times of economic hardship. The role of globalisation, in this case, is not that of having created new social divisions, but that of having heightened the responsiveness- responsibility gap, opening a structural potential for new parties with a populist discourse.

The next chapter looks at the economic and political origins and consequences of the Great

Recession in Greece and Spain in comparative perspective as two illustrative examples.

9 The impact of an exogenous shock such as the Great Recession on politics is largely contingent upon a wide array of variables and factors on the national level and on the interaction dynamics between the context and the actors, which makes the outcomes of the crisis different at the country level. However, there are some common patterns discussed in the remainder of the dissertation that are, for the time being, set apart for the sake of simplicity.

- 16 - Chapter 2. From Boom to Bust: A Comparative Analysis of Greece and Spain under Austerity

Introduction

The economic crisis in Greece has been mirrored by a crisis of the political system: not only has Greece experienced five elections in six years, but in addition a technocratic government was installed for six months and the traditionally bipartisan party system underwent a major transformation. Although the economic crisis has been the deepest in Greece, it has also reshaped political competition in other crisis-ridden countries. In Italy a technocratic government under the leadership of Mario Monti replaced the Berlusconi government in

November 2011 and the political establishment has been undermined by the rise of the populist Five Star Movement in the past few years. In Portugal, the Left Block obtained an unprecedented 10 percent of the vote-share in the 2015 elections and anti-austerity parties were able to form a coalition government.10 Even in Ireland, which is often viewed as a poster child for austerity, the populist party Sinn Fein dramatically gained in popularity and led large protests in 2014 and 2015 against new municipal water charges. Finally, in Spain, large-scale protests have also swept across the country in response to the economic crisis. Two new parties, Podemos and Ciudadanos, have dramatically gained popular support, uprooting the previously stable bipartisan party system.

However, in the context of the economic turmoil in the eurozone, the political transformations in Greece and the other crisis-ridden countries are still ill understood (Hopkin

2015: 163). To analyse these transformations more rigorously, we compare the experiences of living under austerity in Greece and in Spain because of the many similarities between the

10 Austerity is defined as the reduction in government spending during hard economic times.

- 17 - two countries. In both Greece and Spain, the introduction of the euro caused a large inflow of foreign credit, which left both countries vulnerable to external forces when the financial crisis created a sudden stop of liquidity. The economic conditions for the crisis, however, were mediated by the political context. Importantly, both countries suffered from institutional degradation (Royo 2014) in the pre-crisis years and governments and regulators failed to prevent the pre-crisis boom. Because clientelism and corruption allowed the Greek and

Spanish elites to benefit disproportionately from the boom, these same elites had no incentive to lean against the wind and prevent economic imbalances. The windfalls from European integration tempered the political dysfunctionalities in Greece and Spain, but the drying up of foreign resources meant that the state could no longer buy off its citizens with high economic growth. Consequently, a collapse of the political order followed the collapse of the economic order. Widespread dissatisfaction with the political system and a decline in support for traditional parties presented an opportunity for new challengers. Yet, the new challengers that emerged in both countries, and the resulting changes in the party system, were different. We compare Greece and Spain according to the logic of the most similar system design to shed light on these differences.

We argue that the political transformations cannot be understood independently of the social and political dynamics behind the crisis. In other words, we attempt to link the consequences of the crisis to the political context of the countries in question and argue that these political contexts before the crisis not only contributed to the economic crisis, but also structured the political consequences. Moreover, we show that the imposition of austerity from the Troika (the International Monetary Fund [IMF], the European Commission, and the

European Central Bank [ECB]) was a lot harsher and more closely monitored in Greece than in Spain. As a result, the conflict between Greece and its creditors over austerity overshadowed all other political conflicts. This allowed traditional political forces to remain

- 18 - at the heart of the political crisis in Greece: while New Democracy (ND) remained the right- wing pillar of the bipolar party system, Coalition of the Radical Left (SYRIZA), the main challenger party, eventually became an ideological purifier on the left that replaced the former centre-left Panhellenic Socialist Movement (PASOK). In contrast, in Spain the new challenger parties also tapped into the opposition to austerity and dissatisfaction with the functioning of national democracy, but less so toward the European Union (EU). Most importantly, however, these challenger parties have recurrently emphasized their distinctiveness from the previous political order. Although the pro-versus anti-austerity conflict came to dominate politics in Greece, we argue that the division between old and new politics is key for understanding the evolution of political contestation in Spain.

The chapter proceeds in four steps. In the first section we analyse the economic and political context out of which the crisis emerged in Greece and Spain. For this purpose, we review the dominant economic explanation for the crisis and show how the political and economic context contributed to unsustainable economic developments in both countries.

Next, we compare the crisis dynamics in both countries. For this purpose, we analyse the austerity measures that have been implemented in response to the crisis in Greece and Spain and assess the economic consequences. We show that Greece had to endure a lot more austerity than Spain, which is important in order to understand the diverging political consequences. In the third section, we build on this insight to analyse these political consequences of the economic crisis. Distinguishing between a behavioural and attitudinal dimension of the political crisis, we show that there are important similarities in Greece and

Spain but that the consequences of economic meltdown have been conditioned by the domestic political context and the crisis dynamics in both countries. In the final section we outline the implications of our argument.

- 19 - The Origins of the Crisis in Greece and Spain

The Economic Origins of the Crisis

The economic origins of the crisis in Southern Europe have been analysed in-depth. The story usually starts with the adoption of the euro; the euro was criticized at its inception for forcing diverse national economies into a currency union, thus making economic management difficult (e.g., Bayoumi and Eichengreen 1997; for reviews see De Grauwe 2012; Jonung et al. 2009). These problems became obvious at the turn of the century when the German economy was struggling and the Economist (Anon 1999) famously called Germany the “sick man of the Euro,” while other countries in the eurozone were booming. In response to this asymmetry of European business cycles, the ECB conducted a monetary policy that was appropriate neither for the German core nor for the booming periphery (De Grauwe and Ji

2014). It created pressures for divergence within the eurozone and helped to build up large current account imbalances. Large export surpluses in the North were offset by deficits in the

South, which were financed by credit that the North granted to the South (Hall 2012). The result was a large inflow of credit into both Greece and Spain, which increased the liabilities of both countries vis-à-vis non-residents (as represented by the financial account in figure

2.1).

In Greece the government mostly used this inflow of credit as an opportunity to expand government expenditure by borrowing from domestic and foreign banks. This was possible because financial markets had repriced Southern European sovereign bonds and, hence, it became a lot cheaper for the Greek government to issue debt. For several consecutive years the government ran large fiscal deficits, increasing the amount of sovereign debt to 117.5 percent of GDP in 2009. But contrary to popular perceptions, this did not happen in Spain, where government debt remained relatively low. Instead, the ready availability of credit allowed the private sector to go on a spending binge: domestic banks leveraged their balance

- 20 - sheets by borrowing money from Northern European banks to finance a mix of real estate development and consumption in the early 2000s. This led to a huge boom in the Spanish construction sector similar to that in Ireland (Eichengreen 2012), which was also evident in employment figures: more than a fifth of all Greek employees worked in the public sector, but the Spanish construction sector employed nearly every fourth male employee (Fernández-

Villaverde et al. 2013). In other words, unlike the Greek growth model, which was built on an expansion of the public sector, the Spanish growth model was built on an expansion of the construction sector.

Figure 2.1 Financial Account as a pct. of GDP of Germany, Greece, and Spain, 2000–16

● ● ● ● ● ●

● ● 5 ● ● ● ● ●

●

● 0 ● Country ● Germany ● Greece Spain −5 Financial Account

−10

2000 2005 2010 2015 Time Source: IMF 2017.

Nevertheless, the consequences were similar. The large inflow of foreign capital left both countries exposed to an external shock. This external shock came in September 2008, when the financial conglomerate Lehman Brothers collapsed in the US. The global credit crunch led to a flight to safety and a sudden stop of liquidity in Greece and Spain, which resulted in a balance-of-payments crisis (Copelovitch et al. 2016; Wolf 2011). Consequently, the bubble in

- 21 - the construction sector in Spain burst and Greece struggled to finance its deficit on the capital markets. As investors lost trust in the ability of the Greek government to service its debt, the country had to turn to the EU and the IMF for bailouts in 2010, 2012, and 2015. Similarly, eventually the Spanish government could not control the crisis and, on behalf of its banks, it also applied for a bailout in 2012.

The Political Origins of the Economic Crisis in Spain and Greece

The comparison between Spain and Greece shows that the economic origins of the crisis in

Greece were, indeed, systemic. Although the details are different, the economic crisis played out in a similar manner in both countries. With the benefit of hindsight, it is, therefore, easy to argue that there were structural problems with the eurozone (Eichengreen 2012). Yet, this begs the question why politicians and regulators allowed the structural problems to persist in the run-up to the crisis. Much of the existing literature focuses on the economic and financial origins of the crisis and largely ignores the political and social foundation of markets (Polanyi

2001 [1944]). But in order to understand the consequences of the economic meltdown, we need a comprehensive understanding of the crisis, which includes the social and political dynamics behind it.

Matthijs and Blyth (2015) have come a long way in addressing this shortcoming of the literature. They focus on the transnational bargains that are at the heart of the euro and argue effectively that the euro’s lack of embeddedness in proper European institutions is at the heart of the current crisis: not only did the euro lack a proper financial and fiscal union, but it also lacked a political union that could create democratic legitimacy. However, on the domestic levels these institutions existed. For example, the member-states of the eurozone possessed the necessary tools to regulate the banking sector, as Hansen and Gordon (2014) point out.

Why did they not use them?

- 22 - In the absence of effective European institutions, the burden to counteract the growing imbalances in Europe was on domestic institutions. Yet, this burden was placed largely on the shoulders of countries that were least able to do it. First, Greece and Spain both have a weakly institutionalized party system. As democratic latecomers, the party systems were formed when the level of socioeconomic development had already been relatively high (Hopkin

2001b); consequently, the party systems are weakly anchored in civil society. As a response, parties formed a cartel to obtain resources from the state (Katz and Mair 1995, 2009), which they did “either through mass party clientelism and corrupt use of public office to raise money, or through the state funding of parties” (Hopkin 2002: 10). This weakened the patterns of partisanship even further and contributed to the prevalence of informal politics in

Southern Europe.

In Greece two parties dominated the political system before the electoral crisis: ND and PASOK. ND governed Greece’s young democracy in the 1970s, until PASOK came to power with a landslide victory in 1981. Initially PASOK supported radical left-wing policies, including an exit from the European Economic Community, but quickly the party renounced its most extreme positions and shifted toward the centre. In programmatic terms, some differences between ND and PASOK continued to exist, but they paled in comparison to the way these two parties conducted themselves when they were in power: both parties used the state resources to give out favours in return for electoral support (Featherstone 2011). As a result, both parties welcomed the cheap access to credit following eurozone membership. Real government consumption increased by nearly 30 percent from 1995 to 2003, when PASOK was in power, and increased by 17 percent from 2004 to 2009, when ND was in power.

Consequently, it was not in the interest of Greek politicians to prevent the increasing imbalances in the Greek economy in the early 2000s. Rather, they fed off them and used them to entrench their position in society.

- 23 - In Spain the government did not play the same role and the situation is less clear-cut.

In fact, the notion of a clientelistic Southern model should be treated with caution. For instance, Hopkin (2001a) finds little evidence of extensive clientelistic mobilization in Spain, whereas Lyrintzis (Lyrintzis 1984: 101) argues that “clientelism was the major characteristic of Greek political parties”. Still, there is clear evidence of crony capitalism in Spain, much of which was revealed after the outburst of the crisis (Royo 2014: 1574). Contrary to the situation in Greece, where the clientelistic networks were mostly centralized at the national level, the regional and local authorities played a more important role in the case of Spain after the administrative and political decentralization in the early 2000s (Hopkin 2012). Eventually, the political decentralization “led to the emergence of regional, local elites that took over the local and regional institutions. This included the cajas, whose boards were quickly filled with political appointees who used their position for their own personal gain and/or as a clientelist instrument to finance their projects” (Royo 2014: 1574).11 The rapid expansion of the construction sector facilitated by privatization processes increased the opportunities for favouritism, mostly at the regional and local levels.12 Most importantly, the distribution of public contracts for infrastructure projects allowed politicians to generate financial and public resources.

Although initially parties tried to blame individual politicians for this behaviour, the problem was more endemic. Molinas (2012, 2013) argues that an extractive political class had developed in Spain due to institutional arrangements that had been agreed to during the

11. Cajas are Spanish savings banks. Many of them were involved in financing the real estate booms and, as a consequence of the crisis, had to be rescued by government bailouts. 12. This process was even supported by the benefits of membership in the European Monetary Union. Funds from the EU’s Convergence and Cohesion Funds allowed the public sector to finance new railway stations, highways and airports across the country (Hopkin 2012). It is also important to note that while the construction sector also experienced some growth in Greece, the value added of the construction sector paled in comparison to that of Ireland and Spain (see Giavazzi and Spaventa 2011: 213).

- 24 - transition to democracy. For instance, proportional representation and blocked party lists consolidated a weakly institutionalized party system, which the political class used to collude with economic elites to extract resources from society. As in Greece, they formed a powerful set of interests that could resist reforms that would have prevented the increasing economic problems.

As Royo (2014) argues, however, the problem was not one of extractive elites only.

Instead, other vested interests prevented the government from pursuing reforms that would have reduced the dependency of Greece and Spain on foreign credit. Importantly, the benefits of the welfare state across most of Southern Europe are geared toward providing benefits to particular political constituencies (Ferrera 1996, 2005; Rhodes 1997). As Beramendi et al.

(2015b: 395) show, the Southern European welfare states are characterized by weak social investment policies and “particularistic, often residual and regressive social policies of consumption.” In particular, the pension systems are expensive due to high replacement rates, and high employment protection legislation has contributed to a dualized labour market pitting organized insiders against an increasing number of outsiders, who do not benefit from this legislation (Rueda 2007). Before the crisis this created an institutional bias against reform: vested interests had captured the welfare state and were unwilling to forgo these benefits.

The protection of insiders came at the expense of the outsiders, but the economic windfalls from European integration helped to taper the tensions and insulated elites from political pressures for reforms. In other words, there was a vicious cycle: the existing institutional power structure created unsustainable economic growth, which in turn entrenched the institutional power structure further. This power structure not only contributed to the crisis, but also determined how the economic crisis played out in the two countries.

Despite the similarities in the economic and political origins of the crisis, we argue in the next

- 25 - sections that the diverging political outcomes of these two countries are rooted in the way the dynamics of the crisis interplayed with pre-existing structures, thus allowing for a better understanding of the diverging fates of both countries.

Crisis Dynamics in Greece and Spain

The Economic Response to the Crisis in Spain and Greece

Although the symptoms of the crisis were different in both countries, international policymakers prescribed the same medicine. Building on the idea of expansionary fiscal contraction (Alesina and Ardagna 2010; Alesina and Perotti 1997), the creditors insisted on structural reforms combined with a sharp reduction of government spending. With respect to

Spain, this was a fundamental misdiagnosis of the crisis because the Spanish crisis was not caused by government profligacy (De Grauwe 2013). However, creditors viewed Spain through the prism of Greece and condemned both countries to austerity.13

A detailed analysis of the memorandums of understanding (MoU) in Greece and Spain is beyond the scope of this chapter, but it is useful to briefly compare them on a number of dimensions in order to understand their impact. In both countries, the international creditors pushed for austerity and structural reforms. Still, the Greek and Spanish governments had already implemented reforms before they signed the MoUs. Greece came under pressure after the newly elected prime minister, Papandreou, revealed in October 2009 that the budget deficit would be a lot higher than previously predicted. Hence, financial markets became nervous and the credit rating agency Fitch Ratings first downgraded Greece on 8 December

2009 from A– to BBB+ (Oakley 2009).14 Bowing to the pressure of the markets, Greece

13. For an account of the origins of austerity as an idea, see Blyth (2013). 14. This was the first time in a decade that Greece did not have an A rating and reflected the worry in the financial markets about the sustainability of the Greek debt.

- 26 - announced a first set of austerity measurements in February and March 2010 even before the government signed the MoU on 3 May 2010. In the hope of regaining the confidence of investors, these measures included, for instance, an increase of the value-added tax (VAT), a recruitment freeze in the public sector, and cuts to government expenditure. In the absence of a credible commitment from a lender of last resort, however, these measures could not restore the confidence of investors.

Similarly, Spain reacted to the pressure of financial markets before the official bailout.

In response to the 2008 financial crisis, the government had initially pursued a stimulus program; in 2011 the worst seemed over, when the Spanish economy grew again. However, the domestic banks were crippled by many nonperforming loans on their balance sheets, which forced the government to bail them out. In combination with the costs of the initial stimulus program, this increased the government’s deficit and debt burden. In the context of the Greek crisis, international financial markets became increasingly worried about the size of the government’s budget deficit and, starting in May 2010, Fitch downgraded Spain six times until the country lost its A rating in June 2012. Spain introduced austerity measures in 2010 and 2011 to regain the trust of financial markets well before the government agreed to the

MoU on 27 July 2012. In 2011, the Spanish Socialist Workers’ Party (PSOE) and the conservative People’s Party (PP) even amended the Spanish constitution requiring governments to cap deficits at 0.4 percent of GDP from 2020 onward (Tremlett 2011).

The contents of the programs introduced to combat the crisis were also similar in both countries. In Greece, for example, the terms of the first bailout included cuts in public sector pay, additional increases in VAT and other taxes, as well as a limit on public investment and expenditure. In 2010 alone, the aim was to generate savings for a total amount of 2.5 percent of GDP, and by 2014 the fiscal deficit was supposed to be reduced from 13.6 to below 3 percent of GDP (IMF 2010). Similarly, in Spain, in June 2012 Prime Minister Rajoy

- 27 - announced spending cuts and tax increases worth 65 billion euro (Tremlett 2012); as in

Greece, these measures included areas like health care and education. The government also committed itself to increasing taxes and reducing government investment and expenditure, for example by a freeze on the pay of government employees. Moreover, in both countries, austerity measures were combined with structural adjustment programs. The Greek government committed itself to liberalizing closed professions and reducing employment protection; over the years the creditors demanded that the Greek government intensify this program by abandoning the central wage bargaining system, dismissing 15,000 government employees, and introducing a large privatization program. The Spanish government also implemented a labour reform in February 2012 that made it easier to lay off workers and weakened the collective bargaining system. Additionally, the government pursued other reforms that resembled those in Greece, including cuts in social benefits, a rise in the retirement age, and a privatization program to increase government revenue.

Despite these similarities between the austerity programs in Greece and Spain, there are also important differences. First, Spain’s austerity program was shorter. Greece has remained in the spotlight of the crisis ever since October 2009 and it first implemented austerity measures in spring 2010. As the economic recession worsened, Greece was forced to pursue several additional reform programs. In 2012 and 2015 it promised reforms in return for new bailouts; even in between the MoUs, the Greek government implemented additional austerity packages to satisfy the demands of the creditors. Spain also pursued initial reforms in 2010 and 2011, but most measures were announced only in 2012, after the economy had slid back into recession. Most importantly, Spain became the second country in the eurozone to exit the bailout program again in January 2014, while an end to the Greek program is not in sight.

- 28 - Second, international creditors arguably put a lot more pressure on Greece to pursue austerity than on Spain. On the insistence of the German government, the IMF was closely involved in the negotiations of the Greek bail-out and its implementation. Officials from the

IMF joined policymakers from the European Commission and the ECB to form the Troika, and delegates from the three institutions set up camp in Athens and became regular visitors in

Greek ministries. Their presence there increasingly constrained consecutive governments, while they were a lot less influential and visible in Madrid. Moreover, the ECB was reluctant to act as a lender of last resort in response to the Greek crisis in 2010. Only when the crisis had fully reached Spain (and Italy) in 2012 did the ECB change its approach. Fearing an implosion of the euro, ECB president Mario Draghi famously promised to do ‘whatever it takes’ to preserve the euro in July 2012. This change of course was more successful in regaining the confidence of investors than any set of austerity measures could have been: within days, the borrowing costs of Spain and Greece decreased. Hence, Draghi’s announcement came just in time to take some pressure off the Spanish government by reducing the borrowing costs of the crisis countries and by supporting their ailing banking sectors.

Third, the austerity program pursued by the Spanish government was less severe than the program pursued by Greece. Although a comparison between the size of different austerity packages is difficult, the structural balance can be used as a proxy because it excludes the impact of the economic cycle (e.g., through automatic stabilizers) and one-off measures (e.g., the sale of mobile phone licenses) on the government’s budget balance. Using this measurement as a basis for comparison, Greece’s balance improved by 20.3 percentage points from 2009 to 2014, while the structural balance in Spain improved by 8.6 percentage points over the same period. This indicates that Greece pursued a great deal more austerity than Spain—in fact, Greece had to endure the deepest austerity package of all crisis-ridden

- 29 - countries according to this measurement. Although austerity programs created large economic dislocations in both countries, the consequences of the economic crisis were particularly harsh in Greece, reaching an unprecedented severity in modern times.

The Economic Consequences of the Crisis in Spain and Greece

Greece experienced by far the deepest economic downturn of all countries in Europe. It was in recession from 2008 to 2013, when real GDP fell by 26.0 percent. In contrast, Spain had a relatively strong recovery after the global financial meltdown had dragged it into an economic slump from 2008 to 2010. In 2012, however, Spain slipped back into recession, and overall

Spanish real GDP fell by 7.3 percent from 2008 to 2013. This fall in economic activity in

Greece and Spain created similar levels of unemployment in both countries (figure 2.2a) but unemployment before the crisis was higher in Spain than in Greece. In Greece unemployment increased from 7.2 percent in July 2008 to its peak of 27.9 percent in January 2014. In Spain, it increased from 11.2 percent to 25.9 percent over the same period. These levels of unemployment not only were dramatically higher than in Northern Europe, but they were also a lot higher than in the rest of the crisis-ridden South. Still, the true damage of unemployment is evident from the high levels of youth unemployment: in both countries, the share of unemployed under the age of twenty-five tripled between 2007 and 2013 (figure 2.2b), reaching 60 percent in Greece and 56.9 percent in Spain. Consequently, the crisis hit some of the most vulnerable people in society because young labour market participants and low- skilled workers were the first ones to lose their jobs.

However, even many people who retained their jobs were a lot worse off because of the crisis. On average, the wages of employees fell in both Greece and Spain. In Greece this fall was particularly steep: from 2009 to 2014, the average real wage fell by nearly 24 percent in Greece, while the average annual real wage fell by 7 percent in Spain. This contributed to a low level of domestic consumption, but it also increased poverty and social exclusion. In

- 30 - 2014, 36 percent and 29 percent of the entire population was at risk of poverty or social exclusion in Greece and Spain, respectively. Compared to 2007 this figure increased by more than 25 percent in both countries and, consequently, inequality also increased during the crisis according to various measures. While the Gini coefficient rose from 33.4 in 2008 to 34.5 in

2014 in Greece, it rose from 31.9 in 2008 to 34.7 in 2014 in Spain (table 2.1). Similarly, the income of workers at the ninetieth percentile (individuals earning more than the bottom 90% of the population) compared to the earning of workers at the tenth percentile (individuals earning higher than the bottom 10 percent) increased in Greece and Spain by 20 and 18 percent, respectively.

Figure 2.2 Unemployment Rate in Germany, Greece, and Spain, 2000–16

A: Unemployment

20

10 Unemployment Rate Unemployment

Country 2000 2005 2010 2015 Time Germany B: Youth Unemployment Greece Spain 60

40

20 Youth Unemployment Rate Unemployment Youth

2000 2005 2010 2015 Time Source: Eurostat 2017a

- 31 - Hence, the social dislocations of the crisis have been large in both countries and the pains of the economic adjustment were not shared equally across society. The crisis produced a large group of people who were left unemployed, received few social benefits, and suffered from the decreasing quality of public services in health and education (Matthijs 2014). At the same time, austerity policies have not even had the desired effect on government finances. In both countries the government debt in 2015 was significantly larger than at the beginning of the crisis, and the deficits remained outside the 3 percent limit specified by the Stability and

Growth Pact.

Table 2.1 Gini Coefficient in Germany, Greece, and Spain, 2005–16

Year Germany Greece Spain 2005 26.1 33.2 32.2 2006 26.8 34.3 31.9 2007 30.4 34.3 31.9 2008 30.2 33.4 31.9 2009 29.1 33.1 32.9 2010 29.3 32.9 33.5 2011 29.0 33.5 34.0 2012 28.3 34.3 34.2 2013 29.7 34.4 33.7 2014 30.7 34.5 34.7 2015 30.1 34.2 34.6 2016 34.3 34.5 Source: Eurostat 2017b.

The data for Greece and Spain supports this conclusion. Even though the fall in GDP was a lot farther in Greece compared to Spain and any other European country, Greece is no exception when it comes to results of the attempted economic reforms. While Spain has fared better than Greece during the crisis, it is no poster child for austerity either.15 In fact, a comparison with Greece needs to consider not only that the original crisis in Greece was a lot

15. In the wake of the Greek referendum, many commentators have claimed that austerity has worked elsewhere. For example, the Council of Economic Experts in Germany claimed, “Ireland, Portugal, and Spain all successfully completed their program [and that] the economic situation has markedly improved” (Bofinger et al. 2015: 1).

- 32 - deeper than the crisis in Spain, but also that the path of adjustment was a lot more difficult for

Greece due to a more ambitious reform program, as shown above. Moreover, politicians and commentators have consistently and deliberately overstated the positive trajectory of Spain,16 but the average citizen still feels the hangover after the party, as we show below.

The Political Consequences of the Economic Crisis

The Attitudinal Dimension: A Widespread Political Malaise?

It is no surprise that an economic shock with such social repercussions triggered extraordinary political reactions. These reactions were expressed in both attitudinal (lack of confidence in political institutions and the political system) and behavioural (protests, electoral volatility and collapse of party systems) dimensions (Kriesi 2015). For the first dimension, indicators of political trust toward the key representative national and European political institutions, as well as those of the levels of satisfaction with democracy, are particularly revealing. Since the beginning of the economic crisis in September 2008, as marked by the vertical dotted line in our figures, these indicators followed a remarkably similar pattern in both countries. The average of satisfaction with the way democracy works in the national country and in the EU

(figure 2.3) decreased substantially after 2008 in both Greece and Spain, reaching some of the lowest levels recorded since the transition to democracy in both countries.17

16. Fearing punishment from voters and capital markets, politicians have done this deliberately. For instance, Prime Minister Rajoy claimed in a state of the union address, “Spain has passed from being a country on the brink of bankruptcy to a model of recovery that provides an example to other countries in the EU” (Buck 2015). 17. Respondents were asked the following two questions: “On the whole, are you very satisfied, fairly satisfied, not very satisfied or not at all satisfied with the way democracy works in (OUR COUNTRY)?” and “How about the way democracy works in the EU?” We recoded the answer categories, which now range from four to one: 4 = “Very satisfied”; 3 = “Fairly satisfied”; 2 = “Not very satisfied”; 1 = “Not at all satisfied.”

- 33 - Mistrust toward the national government, parliament, and political parties (figure 2.4), followed an almost identical trend.18 Although in both countries mistrust toward political parties had been considerably lower than toward parliament and government before the crisis, we observe that the levels converged to another historical low during the crisis. Compared to the levels of trust toward government and parliament prior to 2008, the decrease is nothing less than spectacular: while more than 50 percent of respondents trusted the national parliament before the crisis, less than 10 percent of respondents trusted the national parliament at the height of the crisis.19

Figure 2.3 Satisfaction with the European Union and National Democracy, 2000–16

Greece Spain

● 0.8 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.6 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●

● ● ● ● ● ● ● ● ● ● 0.4 ● ●

Satisfaction ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.2 ● ● ● ● ● ●

●

2000 2005 2010 2015 2000 2005 2010 2015 Time

● Satisfaction with EU Democracy ● Satisfaction with National Democracy

Source: European Commission 2016.

18. Respondents were asked the following question: “I would like to ask you a question about how much trust you have in certain institutions. For each of the following institutions, please tell me if you tend to trust it or tend not to trust it.” We recoded the answer categories, which now range from zero to one: 1 = “Tend to trust”; 2 = “Tend not to trust.” 19. The recovery of trust in 2015 in the case of Greece is likely due to the victory of SYRIZA in the January legislative elections of 2015. As part of the classic reward-punish mechanisms in democracies, elections tend to momentarily increase trust towards institutions.

- 34 - European political institutions also did not go unpunished: figure 2.5 illustrates that European institutions suffered the largest average drop in trust in Greece and Spain compared to all other institutions. In 2015 there was a slight recovery in these attitudes, but the average levels in both countries remain far from the averages prior to the crisis.

Figure 2.4 Trust in National Institutions, 2000–16

Source: European Commission 2016.

Greece Spain

● ● ● ● ● ● ● 0.6 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.4 ● ● ● ● ● ● ● ● ● ● ● ● ● ●

Trust ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.2 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ●

2005 2010 2015 2005 2010 2015 Time

● Trust in National Government ● Trust in National Parliament ● Trust in Political Parties

Figure 2.5 Trust in European Institutions, 2000–16

Greece Spain

● ● ● ● ● 0.8 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.6 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●

Trust ● ● ● ● ● ● 0.4 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.2 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●

2000 2005 2010 2015 2000 2005 2010 2015 Time

● Trust in European Central Bank ● Trust in European Commission ● Trust in European Parliament

Source: European Commission 2016.

- 35 - Although some studies have shown a strong correlation between the drop of trust in political institutions and macroeconomic indicators, particularly with unemployment (Roth et al.

2013), the widespread political malaise can hardly be attributed solely to economic explanations. Applying the classic literature of economic voting whereby voters punish or reward incumbent governments depending on the economic performance, we would expect that levels of trust toward the incumbent government would have decreased because citizens would blame it for its bad management of the economy. If this were the case, however, trust should recover after citizens are given the opportunity to reward or punish political parties for their performance in government (Key and Cummings 1966a). However, this is not what we observe. Instead, “the cumulated effect of the Great Recession, goes far beyond the short-term punishment of incumbents” (Hernández and Kriesi 2016: 221) and mistrust is indiscriminately projected toward both national and European political institutions, regardless of elections and changes of governments. This suggests that both Greece and Spain suffered a far deeper political malaise than simply discontent with the incumbent’s performance that led to the collapse of their respective party systems. We argue that, beyond the economy, the political consequences of the crisis were related to a feeling of widespread discontent linked to the perception of a democracy without choices (Bosco and Verney 2012), in which mainstream parties were perceived as lacking a differentiated policy program.

The diagnosis is the following: mainstream parties, regardless of their ideology, were strongly constrained in economic policy, leading to a situation where voters could not perceive any significant programmatic differences between them. In other words, the lack of political alternatives that accompanied the austerity programs and reforms triggered widespread disenchantment with the political systems of Greece and Spain. As a result, large sections of society did not feel politically represented and stopped trusting political institutions. More specifically, the neoliberal reforms and welfare cuts implemented by the

- 36 - mainstream left parties (both PASOK and PSOE) under the rhetoric of responsibility resulted in a significant loss of credibility insofar as these parties could not offer an alternative to the right while in government. Although the dynamics of this situation are described in detail later in this section, it is worth noting here that this situation is not the first to take place. The introduction of similar neoliberal-oriented reforms produced an analogous distortion of party systems in Latin America during the era of the Washington Consensus in the 1980 and 1990s

(Mainwaring 2006; Roberts 2013): in countries where the left was in government at the time of implementing cuts, new left populist parties emerged that challenged the true leftist character of the mainstream party. In all these cases, the emergence of the populist left parties was closely linked to the protest arena, the other crucial—and intrinsically intertwined— political dimension of the crisis described in the next section.

The Behavioural Dimension: From the Streets to the Institutions

The attitudinal expressions of the political crisis in both Greece and Spain eventually translated into a behavioural dimension, which manifested itself in both contentious (protests) and conventional (electoral punishment and new parties) politics. In the electoral arena, this took the form of a two-step process in both countries, as shown in tables 2.2 and 2.3. In the first elections during the crisis, the incumbent lost to the mainstream challenger party (ND to

PASOK in 2009 and PSOE to PP in 2011). In a second step, new challenger parties gained a significant share of the votes (2012 and 2015 in Greece and 2015 in Spain).

- 37 - Table 2.2 Spain General Elections Results, 2008–16

Party 2008 2011 2015 2016 PP 39.9 (154) 44.6 (186) 28.7 (123) 33.0 (137) PSOE 43.9 (169) 28.7 (110) 22 (90) 22.6 (85) IU 3.8 (2) 6.9 (11) 3.7 (2) - Podemos - - 20.7 (69) 21.2 (71) Ciudadanos - - 13.9 (40) 13.1 (32) Turnout (%) 73.4 68.9 69.7 66.5 Note: Number of seats in parentheses. IU = United Left. Source: Spanish ministry of interior 2017

Table 2.3 Greece General Elections Results, 2009–15

2012 2015 Party 2009 May June January September ND 33.5 (91) 18.8 (108) 29.6 (129) 27.8 (76) 28.1 (75) PASOK 43.9 (160) 13.2 (41) 12.3 (33) 4.7 (13) 6.3 (17) SYRIZA 4.6 (13) 16.8 (52) 26.9 (71) 36.3 (149) 35.5 (145) KKE 7.5 (21) 8.5 (26) 4.5 (12) 5.5 (15) 5.5 (15) XA 0.3 (0) 7 (21) 6.9 (18) 6.3 (17) 7 (18) ANEL - 10.6 (33) 7.5 (20) 4.7 (13) 3.7 (10) Turnout (%) 70.9 65.1 62.5 63.6 56.6 Note: Number of seats in parentheses. KKE = Communist Party of Greece; XA = Golden Dawn; ANEL = Independent Greeks. Source: Greek ministry of interior 2017

However, while this two-step electoral process was similar in the two countries, there are significant differences in the protest dimension of the crisis that are worth stressing. In Spain, the mass demonstrations that would give the name to the 15-M movement—and that would later trigger similar protest in the United States under the “We are the 99 percent” banner

(Castañeda 2012)—were initially convened by a small online platform called Democracia

Real Ya! (Real Democracy Now!) in May 2011. With a simple, direct and nonpartisan message, Spanish people mobilized both inside and outside Spain. “We want more democracy!,” “We are not commodities of bankers and politicians!,” or “They do not represent us!” were some of the slogans that were coined by the movement. The so-called

Indignados movement could connect with large sections of the population despite (or perhaps

- 38 - because of) the fact that they refused any involvement with traditionally politicized organizations such as labour unions or youth organizations of parties. More than just anti- austerity protests, the movement had to do with providing “time and space for every participant to express themselves and take an active part in the camp and movement organization, notably through long group discussions and the creation of commissions and working groups around specific issues” (Kaldor and Selchow 2015: 206).

The causes of contention that sparked the movement were rooted in the perception that an extractive and corrupt elite had governed unaccountably and made the citizens pay for a crisis that they had not caused. Whereas austerity was an important mobilizing factor for the protests, the movement also repeatedly identified the political structure of the country as one of the main problems (Calvo et al. 2011). The idea of new politics grew as an alternative to old or traditional politics, which, to the movement, was represented by the extractive economic and political elites. At the core of this idea were participatory mechanisms and claims for more direct democracy, with the use of the internet and social media being its main driver (Hughes 2011). The Indignados movement was, therefore, not solely an anti-austerity protest movement, but also a much wider expression of the dissatisfaction with the Spanish political system. In fact, it was the inception of this new conflict in Spanish politics that would later be expressed with the emergence of new parties like Podemos and even

Ciudadanos, as explained below.

Although the so-called Greek Indignados also played a role in the anti-austerity protests from 2010 to 2012, they were neither the only actor nor the most important one. In fact, anarchists, social justice groups, educational groups, youth organizations, but especially political parties and labour unions (from the private and public sector), were equally if not more important—depending on the individual protest—in participating and organizing the protests. For instance, the two large trade union confederations, the General Confederation of

- 39 - Greek Workers (GSEE) and the Civil Servants’ Confederation (ADEDY), as well as the youth organizations of the Communist Party of Greece (KKE), SYRIZA, and ANTARSYA, played a key role in the organization of several major protests in the country (Kanellopoulos and Kostopoulos 2014). One of the main differences between the way in which anti-austerity protests were organized and developed in Spain and Greece—which is particularly important to understand the political consequences and the characteristics of new parties—is precisely that the Indignados in Spain “mainly involved young and highly educated people not connected with ‘old’ political actors like trade unions and political parties” (Karyotis and

Rüdig 2015: 496). This difference is central for explaining how these movements translated into the conventional political arena and the nature of the new challengers.

The New Challengers: A Populist Left?

The strong dissatisfaction with traditional parties and the mobilization in the contentious arena opened an exceptional opportunity for new political entrepreneurs to capitalize on the political crisis. Social democratic parties in both countries faced an irreconcilable dilemma that would ease the emergence of new parties on the left. This dilemma, framed by Mair

(2009) as responsibility versus responsiveness, describes the situation confronted by both

PSOE and PASOK. Both had to introduce measures contrary to their ideology—whether through changes in the constitution or through the implementation of cuts on welfare—in order to behave responsibly and satisfy the demands of the international creditors (Bremer

2018). This came at the cost of responsiveness to the ideological foundations of these parties.

In Greece this implied the near disappearance of PASOK, going from 43.9 percent of the vote in 2009 to 4.7 percent in 2015. By implementing austerity after its election in 2009, the party not only had committed an ideological somersault, but by depleting the resources available to distribute, it had also undermined the clientelistic linkages that the party depended on (Afonso et al. 2014). SYRIZA benefited the most from this decline of PASOK, going from 4.6 percent

- 40 - in 2009 to winning both elections in 2015 with over 35 percent of the votes. Not only did most of SYRIZA’s votes came from PASOK, but in addition even some politicians changed affiliation to the new party. As PASOK lost the credibility to stand up to the European creditors and to implement a leftist program, SYRIZA became PASOK’s ideological purifier, a type of challenger party that is successful when established parties are forced to change their position with regard to salient lines of conflict (Lucardie 2000). With a radical leftist and anti- austerity agenda, the party promised to be responsive toward the electorate by standing up to the international creditors’ demands and putting an end to austerity, while at the same time combating the domestic oligarchy. Thus, SYRIZA confronted both the domestic and the

European elites, dubbed as the internal and the external Troika.

Similarly, in Spain the initial challenger party that emerged from the Indignados movement was Podemos, which became the third-largest party in the 2015 election winning nearly 21 percent of the vote. Although both SYRIZA and Podemos have been labelled as parties of the new populist left in Southern Europe, and despite the visible links between both parties,20 there are significant differences that suggest that they are two different animals.

Most visibly, SYRIZA was presented as a clear alternative from the left and had a long political tradition, despite its weak parliamentary representation in the past. In contrast,

Podemos was a completely new party, which, in line with the ideas of inclusiveness of the 15-

M movement, shed traditional left-right ideological labels. A more analogous situation would have been for the United Left (IU) in Spain, a traditional leftist party with a clear anti- austerity message, to “purify” PSOE, just as SYRIZA did with PASOK. Why did this not happen? Why did it take a new party such as Podemos to capitalize on the emerging protest vote?

20. On 22 January 2015, for instance, Alexis Tsipras symbolically invited Pablo Iglesias, the leader of Podemos, to the end- of-campaign meeting before the elections.

- 41 - The electoral arenas of politics in Greece and Spain were closely influenced by developments in the arena of contentious politics. Whereas in Greece the anti-austerity discourse was one of the key issues in all the demonstrations (i.e., demands that the rich are taxed, social welfare provided, ceasing of privatizations, etc.), which was the political territory of the left, this was not so clear in the case of Spain. As explained above, the

Indignados emerged as an all-encompassing movement that persistently claimed to be inclusive to all citizens regardless of their ideology, pursuing the idea of new politics. Thus, the issue of new politics, including corruption and the opposition to the domestic political system, became salient in Spain in both the contentious and conventional arena of politics, while the intervention of external European actors in domestic politics did not play a determining role as much in Spain as it did in Greece. In fact, Calvo, Gómez-Pastrana, and

Mena (2011) show that the most important objectives of the 15-M movement were the fight against corruption and the reform of the electoral system. As explained in the previous section, this was because the intervention of these external actors became much more pronounced in the case of Greece. Therefore, Europe and the enforcement of austerity played a smaller role in reshaping the Spanish political system than that of Greece (see also Chapter

3).

The greater importance of domestic renewal over European anti-austerity issues also explains the strong rise of a new politics right-wing party in Spain, Ciudadanos, which has gained rapid electoral support since the last regional elections in 2015.21 Despite being a pro- austerity party, it entered the conventional political arena as the right-wing equivalent of

Podemos, which also challenged the domestic elites but had little to say about Europe.

Although Ciudadanos self-identifies as a liberal party with a very different program from that

21. Although there is a similar new party in Greece, To Potami (The River), it obtained only 4 percent of the votes in September 2015 and 6 percent in January 2015.

- 42 - of Podemos, their voters share a set of characteristics that place both parties in direct competition with each other on this new dimension of conflict despite being on different sides of the ideological spectrum. Compared to the voters of PP and PSOE, voters for both parties are more concerned with the political situation and are more likely to be young and from urban areas (see Chapter 4), indicating a new dimension of political conflict in Spain.

This new dimension of conflict in Spain also explains why it would take a new party such as Podemos to capitalize on the emerging protest vote. Although the traditional left party

IU had long campaigned against austerity, it did not correspond with the idea conveyed in the contentious political arena about the necessity of new politics and was, thus, not able to capitalize on the emerging protest vote. Being perceived as part of the old politics weighted heavily among young and urban voters, who were demanding a change in political culture.

Conversely, the dominance of the new dimension of conflict in Spanish politics can also account for failure of Podemos to purify PSOE. While the difference in the depth of austerity

(and, thus, the degree of ideological betrayal) that the social democratic parties were forced to implement is also important for explaining the extent that voters and politicians switched, it is only one part of the story. Importantly, many old political forces, including the trade unions, remained loyal to PSOE because they also felt threatened by the rise of the new challenger parties in Spain. In contrast, one of the important characteristics of the Greek case was

SYRIZA’s capacity to mobilize the mass demonstrations with the participation of major unions. Thus, while in Greece old actors (i.e., major unions) and new actors (i.e., Greek

Indignados) joined forces under the same banner, the division between old and new actors weighted heavily and defined the dividing lines of Spanish politics during the crisis.

The political developments of Greece and Spain on the eve of the Great Recession resemble each other, but there are important differences that explain the diverging patterns in the shaping of the new structure of conflict. The restructuring of the party systems in both

- 43 - countries was different because the main challenger in each country responded to different patterns in the contentious arena, which led to a different institutionalization of conflicts.22

Due to the intensity of the crisis, the conflict about austerity dominated in Greece, while a conflict between old and new politics became more prominent in Spain. In other words, the domestic dimension prevailed in Spain, whereas in Greece the European dimension became more important (see Chapter 3: 63). Naturally, these new conflicts are deeply intertwined with other idiosyncratic conflicts in each country. For instance, the rise and success of Ciudadanos cannot be understood without the territorial conflict in Spain because the issue of independence for Catalonia is key to Ciudadanos’s program (Teruel and Barrio 2015).

Although a full discussion of these issues is beyond the scope of this chapter, it is important to note that the domestic political culture of countries is essential for understanding the new political conflicts in Greece and Spain.

Conclusion

The response to the economic crisis in Europe has been driven by neoclassical economic theory. Policymakers have attempted to turn the South into the North in order to create the optimal currency area that Europe has never been. However, in this chapter we have argued that the introduction of the euro is not exclusively to blame for the crisis and the resulting imbalances in Europe. Instead, prior political dynamics also played an important role. The cases of Spain and Greece are tellingly similar in that respect: both countries suffered a period of institutional degradation (Royo 2014), where governments and regulators failed to prevent the pre-crisis boom. Clientelism and corruption, in its different levels and forms, allowed the

22. Although the leaders of Podemos speak of overcoming the left–right spectrum in order to attract voters regardless of their ideological position, studies have shown that citizens persistently place the party on the very left of the ideological scale (Fernández-Albertos 2015a).

- 44 - Greek and Spanish elites to extract resources and benefit from the pre-crisis boom disproportionality. In combination with the support from entrenched interests that benefited from the dualized welfare system, this disincentivized regulation that would have controlled the excesses in the public sector (in the case of Greece) and the private sector (in the case of

Spain). In the pre-crisis era the economic windfalls of European were large enough to shield the political system from domestic opposition, but the political dysfunctionalities of Greece and Spain were exposed in the past few years. In the wake of the crisis, the trust toward national and European political institutions dropped dramatically as citizens became increasingly dissatisfied with the political system. This dissatisfaction was rooted not only in the declining performance of the economy, but also in the political context—that is, the clarity of responsibility and the availability of alternatives. The result was an attitudinal crisis that was similar in Greece and Spain, but that evolved into two distinct behavioural patterns in the two countries.

In Greece the austerity programs were so harsh and comprehensive that they overshadowed every other political issue. Thus, the international creditors, personified by

German politicians and bureaucrats of the Troika, were quickly regarded as the main culprit for the economic situation in Greece and, as a result, the main conflict in Greek politics became one that posed domestic political actors against international actors (and their domestic allies). In Spain, in contrast, the political crisis played out differently. The austerity program in Spain not only was less deep, but also the European institutions did not have the same (perceived) involvement in its design. While the austerity policies were also politically divisive, the main political conflict in Spain became an internal one: political forces, who stood for old or traditional politics, were attacked by new protest movements and political parties, which were supporting a form of new politics cantered around claims for participatory mechanisms and more direct democracy.

- 45 - The new political parties that have emerged out of these conflicts are representative of these patterns and, despite their common stance against austerity, SYRIZA and Podemos are different political animals. These differences are best explained by the developments in the contentious arena of both countries, which created the political opportunities for the new challengers. Whereas in Greece it was the traditional left that capitalized on the emerging discontent, it took a new party that shunned ideological labels in Spain to do so, at least initially. Therefore, the behavioural consequences of the crisis were different because of a combination of pre-existing domestic grievances and new conflicts that were brought about by an exceptional economic and political situation.

Our comparison of Greece and Spain shows what an encompassing effect the economic crisis and the austerity policies had on the Greek political system and society.

Although other European countries like Spain also suffered heavily during crisis, the shackles of austerity in Greece were unprecedented. The Greek political system underwent radical transformations. Importantly, these transformations are far from over. The transformations of the Greek political system have been that of a punctuated equilibrium: it is once again dominated by two political parties: the conservative ND on the right and the governing party

SYRIZA on the left. In 2015 SYRIZA signed a third MoU including further austerity and reforms without the guarantee of much-needed debt relief. Because SYRIZA is struggling to live up to its political promises to end the austerity regime, the political situation is still subject to further changes. Overall, the political consequences of the crisis in Greece have already been colossal, but they might have only been the prelude of what is still to come.

- 46 - Chapter 3. Old versus new politics: The political spaces in Southern Europe in times of crisis

Introduction

The economies in the European South have been particularly hard hit by the fallout of the

Great Recession and the subsequent Euro crisis. Economic indicators show that most countries in North-western and Central-Eastern Europe recovered fairly quickly after the first storm of the Great Recession which travelled to Europe after the collapse of the US-

American bank Lehman brothers in fall 2008. By contrast, the Southern European economies were caught in a spiral of stagnation, rising unemployment rates (most dramatic among the young) and public debt for the last 8 years (see Conti et al. 2018). Given the strong economic and political interdependence in Europe (especially within the Eurozone), the political processes to cope with the economic crisis involved hard bargaining among European states and supra- and international institutions. On the European Union (EU) level, these bargaining processes have led to conflicts between governments representing their national interest as

“debtor” (Southern European and Ireland) versus “creditor” (North-western European) countries. On the domestic level, especially the so-called “debtor countries” saw major changes of their party systems as highlighted by unprecedented electoral volatility and the rise of new challengers from the left and right (Hernández and Kriesi 2016; Hobolt and Tilley

2016).

In this article, we focus on the way these events have affected the emerging political spaces in four Southern European countries—Greece, Italy, Portugal, and Spain— since the onset of the Euro crisis. That is, we focus on the supply side of domestic party competition by studying the key issue divides that structure the competition and the actors’ configuration within the emerging political spaces. Our contention is that with the Great Recession and the

- 47 - subsequent Euro crisis, the configurations of the political spaces in Southern Europe are changing. Before the crisis, the Southern European partisan spaces remained essentially bipolar with cultural and economic issues amalgamated in one single left–right dimension

(Polk and Rovny 2015) and parties weakly rooted in the cleavage structure (Gunther 2005).

This contrasts to the pattern observed in North-western Europe where, driven by consecutive mobilization of the new left and populist right, the divides over economic and cultural issues are rather orthogonal to each other (e.g., Hooghe and Marks 2017; Kriesi et al. 2008, 2012).

However, given long-term trends and the way the crisis played out in the domestic arenas of the countries under scrutiny, we do not expect the same pattern to emerge as in

North-western Europe. By contrast, to understand the political changes underway in Southern

Europe, we argue for the need to consider that these countries simultaneously face an economic and a political crisis which both have strong domestic and European components

(see also della Porta 2015). This should give rise to salient conflicts over both austerity and political renewal. Moreover, the key driving forces of change are social movements and political parties that forcefully combine opposition to austerity and “old politics”. Thus, we expect an alignment of the divides over economic and political issues in the political spaces and that this alignment is even stronger when the mainstream left is in opposition. In government, the mainstream left is forced to implement austerity that leads to a “neoliberal convergence” of the mainstream parties (Roberts 2013), whereas it attempts to capitalize on claims for political renewal and anti-austerity when in opposition. To put it differently, once the mainstream left in opposition opts for an “accommodative” strategy1 regarding the

1 Meguid (2005) differentiates between three major strategies of mainstream parties vis-a`-vis new challengers and their claims: dismissive (i.e. neglect the issue), adversarial (i.e. oppose the issue position of the challenger), or accommodative (i.e. adopt the issue position of the challenger).

- 48 - challengers’ demands, the resulting political space should be characterized by strongly aligned— that is, not orthogonal—divides over economic and political issues.

This article is structured as follows. To prepare our argument, we first discuss the economic and political crises in the four countries. Next, we turn to the challengers in the party systems and their differing emphasis on the failures of domestic and/or European elites.

This leads to more specific expectations on the emerging structure of conflict. Third, we introduce design and methods. We focus on the nine national elections that took place in the four countries from 2011 to 2015. By doing so, we move beyond an analysis of single

“extraordinary” elections—like the Greek election in May 2012 or the Italian election in 2013

(e.g., Conti and Memoli 2015; Katsanidou and Otjes 2015a)—in our quest to assess the consequences of the crisis for the emerging conflict structure. Empirically, we follow the research strategy of Kriesi et al. (2008, 2012) and rely on a relational content analysis of media data. This original new data set allows studying public debates as they unfold during election campaigns. In the conclusions, we summarize the key findings, situate them in the broader comparative picture of European party politics, and point to limitations of our study and directions for future research.

Overall, the findings support our expectations. Putting the accent on similarities, we find that both economic and political issues are key to understanding the structure of political conflict in Southern Europe in times of crises. All countries saw significant conflicts over austerity and old versus new politics, and there is a strong association between opposition to domestic austerity and calls for democratic renewal. This pattern emerges everywhere, although its concrete manifestation varies across countries and over time depending on, among others, the government participation of the mainstream left.

- 49 - Conflicts in times of economic and political crises

From our point of view, it is useful to combine structural and strategic approaches to party competition as they tend to complement each other (for the two approaches, see De Vries and

Marks 2012: 187–190). As argued by the structural perspective, processes of social change create structurally defined potentials that are, as argued by the strategic approach, mobilized by strategic political actors. Of course, an approach that views party competition as ultimately rooted in structural conflicts begs the question of which social conflicts have the capacity to restructure the partisan space. Given the devastating effect of the crisis on the

Southern European economies, we expect political conflict to focus on economics. However, for understanding the political consequences of the economic crisis, it is crucial to keep in mind that Europe has developed into a multilevel governance structure and that, given the close economic and political integration of EU member states, the economic crisis in Europe has developed into the “Euro crisis” (Copelovitch et al. 2016). This crisis has been mainly driven by economic imbalances between the Northwestern and Southern members of the

Eurozone (e.g., Lane 2012; Scharpf 2011). The governments of the weaker, Southern

European economies, in particular, were unable to cope with the crisis, and the EMU governance structures revealed their structural weakness (e.g., Eichengreen 2012).

As stated, the ensuing crisis management involved above all the EU’s intergovernmental channel, and the European governments represented their national interest as “debtor” (Southern European and Ireland) or “creditor” (North-western European) nations in this bargaining process—whatever their partisan composition (e.g., Laffan 2016). Under the pressure from the EU (represented by the “troika”), the national governments adopted austerity policies that were harsh for large parts of society. The model case is Greece, where the troika intervened most manifestly and with most dramatic consequences for the country’s economy (see Chapter 2). However, under the impact of the crisis, the other three countries

- 50 - under scrutiny became the object of supranational interventions as well. Portugal was bailed out by the International Monetary Fund/European Financial Stability Facility in spring 2011.

Spain accepted a bailout of its banks by the European Stability Mechanism in summer 2012.

Italy, even if not formally bailed out, became the object of “implicit conditionality” when it was hit by the financial storm in summer 2011.

Table 3.1 (Expected) Conflicts in Southern Europe

Economic Political

Domestic austerity Domestic democratic renewal Domestic Welfare, economic liberalism Democratic reforms, corruption

European austerity European integration European Euro, bailout and its conditions European integration, deepening

As a result, the governments’ manoeuvring space in macro-economic policy making was severely restricted, with significant political consequences at the domestic level. In Italy, the imposed austerity measures brought down the right-wing Berlusconi government and ushered in the technocratic Monti government that was supported by all three major parties (including the mainstream left Democratic Party [PD]). In Greece, Portugal, and Spain, the imposed austerity policies initially had to be executed by governments from the left. These governments were not able to adopt the reforms they had originally promised and were forced to take measures that contradicted the policy positions for which they were known in the past. In line with the Latin-American experience (e.g., Lupu 2014; Roberts 2013), the neoliberal convergence of major parties imposed by forces external to national party competition led to party brand dilution, to a decline of partisanship and, eventually, to a process of de-alignment—and, in the case of Greece, to the collapse of the incumbent

Socialist party (PASOK) and even the entire party system (e.g., Verney 2014).

- 51 - However, the challenging of the mainstream parties under the impact of the crisis had also domestic origins. Structural problems, policy errors, and misconceptions pre-dated the

Euro crisis and left the Southern European countries particularly ill-prepared to respond to the crisis. All four countries are characterized by weak state capacity (Beramendi et al.

2015b: 13), and even this capacity has been systematically undercut by clientelistic practices and political corruption. Greece certainly has been the worst offender with both major parties establishing a system of party patronage (Pappas and O’Malley 2014), while the Portuguese parties have been comparatively less clientelistic (Afonso et al. 2014: 6). As Royo (2014) argues for Spain, we cannot understand the real estate bubble, the loss of competitiveness or the financial crisis without taking into account what he calls the “institutional degeneration” in Spanish politics. However, as he also argues, the problem is both the extractive behaviour of elites and that civil society tolerated such behaviour. Only when the crisis exposed an unsustainable economic model, the public was outraged by the actions of its elites.

The increased perception of corruption was coupled with growing distrust toward both domestic and European political institutions (Muro and Vidal 2016, see also Chapter 2).

An overall sense of frustration with the political elites captivated a large share of the population in Southern Europe and soon translated into activity in the protest arena—signs of what Mainwaring (2006) calls the attitudinal and behavioural dimensions of a political crisis.

While the protests in the four countries differed in timing, size, and organizational sponsors, they emphasized similar claims against austerity measures and the malfunctioning of

(representative) democracy (see Altiparmakis and Lorenzini 2018; della Porta 2015).

What does this mean for the structuration of the partisan political space? First of all, it is important to keep in mind that there are two overlapping conflicts—the conflict with the domestic elites and the conflict with the European elites. Second, each one of these conflicts has a political and an economic component. To illustrate this, Table 3.1 shows a fourfold

- 52 - table with the four combinations of conflicts and the associated issues. The domestic conflict is about austerity policies (an economic issue) and corruption and democratic renewal (a political issue). The supranational conflict, where it is present, is obviously about austerity, too, but it is also about the defence of the nation state, about national pride and humiliation, and the democratic deficit at the European level. To sum up, we contend that the political spaces in Southern Europe reflect the economic and the political crises that have unfolded in these countries since the outbreak of the Great Recession in 2008. Thus, we formulate the following hypothesis:

Hypothesis 1: Under the impact of a twofold crisis, conflicts over economic (left vs. right) and political issues (new vs. old politics) are structuring the political spaces in Southern Europe (double-crises hypothesis).

The rise of (different) challenger parties in Southern Europe

New challenger parties are the driving forces of the transformation of the political space given the programmatic inflexibility of mainstream parties and their tendency to stick to the dominant dimensions of conflict (for a recent summary, see Hooghe and Marks 2017). They

“translate” the latent potentials into manifest conflicts. Therefore, we briefly introduce the electorally most successful new challengers in the four countries as this allows refining our expectations.

Once the economic crisis struck, the electorate reacted in a two-stage process in the electoral arena. In a first stage, incumbents were punished as predicted by the economic voting literature. The punishment was exemplary, but it was rather conventional as voters turned to the mainstream opposition party. Only in the second stage, the voters typically sanctioned all the mainstream parties and massively turned to challenger parties (e.g.,

Hernández and Kriesi 2016; Hobolt and Tilley 2016). Portugal is the only country among the four where there is no marked second stage in sight yet, even if in the 2015 elections an

- 53 - unprecedented 18.5% of the vote went to parties left of the socialist PS— the Left Bloc BE gaining the most.

In Greece, challengers from both the left and the right rose to prominence in the 2012 elections (e.g., Gemenis and Nezi 2015; Katsanidou and Otjes 2015a; Verney 2014). On the right, the period saw the rise of different challengers from the extreme right Golden Dawn, the populist right Independent Greeks (ANEL) to the new center-right party The River (To

Potami). However, by far electorally most successful proved to be the Coalition of the

Radical Left (Syriza), becoming the largest party in the January 2015 election. Syriza was a left-wing opposition party that had been founded in 2004 and that established close links with the extraordinary mobilization of Greek society in the streets. Syriza, besides holding a strong anti-austerity (and initially anti-bailout) position, opposed both the domestic and the

European elites. As observed by Stavrakakis and Katsambekis (2014), Syriza’s discourse became clearly populist by painting a sharp antagonism between “us” (“the people”) and

“them” (the “elite”). Tsipras, Syriza’s leader, used the phrase of “external troika–internal troika.” In this way, the three-party coalition government of New Democracy (ND), PASOK, and the Democratic Left (DIMAR) was effectively equated with the country’s international emergency lenders. While the opposition to both austerity and the domestic and European elites unites the Greek challengers from the right and left, Stavrakakis and Katsambekis

(2014: 132) point out that Syriza’s populism tends to be inclusionary. They highlight that

Syriza has been “one of the most consistent advocates of the immigrants’ equal rights and their full inclusion in Greek society” (p. 132). Syriza also supports common claims of the new social movements and “new left” parties of the 1970s such as gender equality and LGTB rights.

In Spain, too, challengers rose both on the left (Podemos) and the centre-right

(Ciudadanos). Although the electoral results do not differ as much as in the Greek case, the

- 54 - new left challenger, Podemos, is so far also more successful than its centre-right rival (e.g.,

Ramiro and Gomez 2017). Podemos grew out of the social movements of the squares. It was founded early in 2014 and had its first success in the European elections in March 2014. In the 2015 general elections, it obtained 20.7% of the votes. Podemos, as opposed to Syriza, is more focused on the national elite and less anti-European. In unmistakably populist manner, the party claims to defend the “social majority” (“mayoría social”), the “power of the people”

(“poder popular”) or simply the “citizenship” (“ciudadania”) against the degenerated “casta” of politicians (e.g., Ramiro and Gomez, 2016). The focus on domestic elites is equally if not more pronounced with Ciudadanos than with Podemos. Ciudadanos, formed in summer 2006, it grew out of civil society in opposition to nationalism in Catalonia (Rodríguez Teruel and

Barrio 2016). Its ideas concern above all domestic reforms (of the party system, the legal system, public administration, and internal decentralization) as well as economic and social policy. Both parties are not anti-European. Despite its anti-austerity positions, the program of

Podemos is conscious of the fact that some of its demands can only be implemented in the

European context. While at variance with the old Communist left, this is in line with the tradition of the Spanish mainstream left for whom Europe provided relief and shelter from the fascist past (see Díez-Medrano 2003). Thus, Podemos proposes a new “Carta

Democrática Europea,” which intends to provide citizens with greater participation in

European policy-making and ethnic minorities, such as the Catalans, with self-determination.

Although Podemos and Ciudadanos differ in ideological terms, their voters share a set of socio-demographic (younger, more educated, urban, and socio- cultural professionals) and attitudinal traits that point to the fact that both parties represent those particularly concerned with the political crisis in Spain (Chapter 5).

In Italy, the new challenger, the five star movement (M5S), is harder to situate, and its claim to be neither left nor right is also reflected in its electoral base (e.g., Bobba and

- 55 - McDonnell 2015: 172; Ceccarini and Bordignon 2016: 143).2 It began a startling rise in opinion polls from below 6% in April 2012 and obtained no less than 25.6% of the vote in the national election in spring 2013. In its program, the party casts the citizens of Italy as the victims of a system captured by corrupt and incapable elites at the national and supranational level. The blame is firmly laid on the shoulders of Italy’s entire ruling class (comprising all existing parties, the media, and business leaders) and European elites, to whom the party attributes the malfunctioning of democracy and the economy’s decline (e.g., Bobba and

McDonnell 2015). To remedy the situation, M5S proposes to overturn the system, removing the current elites and restoring power to the Italian citizen by online direct democracy. As many of its supporters have defected from the right and given some anti-immigration claims in its program, the Italian M5S fits least into the new left category (Ceccarini and Bordignon,

2016: 154). However, this challenger, too, has some features in common with the new social movements and the parties they spawned. As Biorcio (2014: 37, 2015: 121) has observed, especially with its calls for a participatory and deliberative democracy, this movement recalls the German Greens 30 years ago.

Based on the programmatic profile of the challengers, we can specify our expectation further. Overall, the challengers in Southern Europe—like their counterparts in the North— raise specific policy demands, but they also pose a challenge to the established system of interest intermediation more generally (Hutter 2014). To a certain extent, they combine the new left’s call for participatory democracy and its reliance on protest politics with the radical right’s populist appeal. While the challengers differ to a certain extent in their program and the focus on the European dimension, they all share a strong focus on both the economic and political questions at stake (again, see Table 3.1). Moreover, in all countries the paradigmatic

2 As a matter of fact, the Spanish Podemos, although clearly to the left, initially adopted a similar strategy as M5S claiming to be neither left nor right in order to capture the center of the political spectrum.

- 56 - challengers from the populist radical right prevalent in North-western Europe are not among the electorally most successful challengers. The more central role of challengers from the left seems due to the discrediting of both the populist radical right and the mainstream left, and it mirrors the pre-crisis situation as leftist challengers have traditionally been stronger in the

South of Europe than in the Northwest (Hooghe and Marks 2017: 15). The populist radical right was largely discredited by the authoritarian legacy, the fact that immigration has hardly been an issue in these emigration countries until recently, and that European integration was essentially seen in positive terms.3 Moreover, the programmatic dealignment of the mainstream left resulting from the “forced” implementation of a neoliberal program opened up a niche for challengers to capitalize on both conflicts through combining an anti-austerity message with a need for regenerating the political system in general. Thus, we expect that the relationship between the conflicts shown in Table 3.1 is not orthogonal to each other. Rather, due to the integration of opposition to austerity and calls for democratic renewal on the left side of the space, we expect an alignment of the divides over economic and political issues.

This alignment of the economic and political divides in the political space should be even stronger when the mainstream left is in opposition. As argued before, the mainstream left implemented austerity when in government which led to a certain convergence of mainstream left and right on economic issues and opened up a niche for the emergence of challengers with an anti-austerity and anti- old politics agenda. By contrast, we expect that the mainstream left attempts to capitalize on claims for both political renewal and anti- austerity when in opposition. Such an accommodative strategy resembles responses of the social democrats to the rise of Green parties (Kitschelt 1994) and of the mainstream right to the rise of populist radical right parties (Akkerman 2015). If the mainstream left in opposition

3 Italy being the most deviant case (see Gattinara 2016)

- 57 - opts for such a strategy, the resulting configuration resembles a reconfigured bipolarity between the mainstream right and the rest. Ultimately, this should be reflected in strongly aligned—that is, not orthogonal—divides over economic and political issues. To sum up, we formalize the discussion in two hypotheses:

Hypothesis 2: The economic and political divides are closely aligned with each other—that is, not orthogonal—in the political spaces in Southern Europe given that the driving forces of change are challenger parties that combine opposition to austerity and “old politics” (alignment hypothesis).

Hypothesis 3: The alignment of the economic and politic divides is more pronounced when the mainstream left in opposition adopts the demands of the challenger parties as compared to the situation when it is in government (accommodation hypothesis).

Design and methods

In the following section, we briefly introduce the design and methods (a more detailed methodological discussion including a note on our strategy of data analysis can be found in the Appendix). We adopt a most different systems design as we are most interested in carving out what is similar in the structure of the political spaces in the four countries taking into account that they differ in many respects (from the electoral systems to the strength of other traditional cleavages, such as the centre-periphery cleavage). Moreover, the previous sections have already pointed to crucial differences regarding both the latent potentials and the actors who articulate them, that is, the extent to which the two crises affected the countries under scrutiny and the types of new party challengers that have been emerging. The campaigns covered by the analysis took place when the Euro crisis was in full swing. More precisely, we focus on the four Greek elections in May and June 2012 and in January and September 2015; the Italian election in 2013; and the Portuguese and Spanish elections in 2011 and 2015.

- 58 - To analyse party competition as it unfolds during electoral campaigns, we follow the strategy of Kriesi et al. (2008; 2012) and make use of a relational content analysis of newspaper articles. This strategy allows us to study the publicly visible contestation among the political parties during the campaigns and map the issues of the day onto broader issue dimensions to test our claims about the structure of the political space. The analysis is based on the coding of two newspapers per country. To reduce ideological biases, we selected two quality newspapers per country, one each from the centre-left and the centre-right. We then coded a sample of the selected articles using core sentence analysis (CSA). Following this type of relational content analysis, each grammatical sentence of an article is reduced to its most basic “core sentence(s)” structure, which contain(s) only the subject, the object, and the direction of the relationship between the two. For the following analysis, we rely on all relations between party-affiliated actors as subject and any political issue as object. Overall, the analysis is based on around 11,000 such actor-issue statements (see Table 1A of the

Appendix).

A crucial step in our analysis is aggregating the detailed issues that were coded into a set of broader categories. The aggregation was guided by two considerations: (a) to capture the general conflicts described in the previous sections and (b) to compare our results to previous findings of Kriesi et al. (2008; 2012). More specifically, we regrouped the detailed issues into 18 categories (see Table 2A of the Appendix). For the analysis, the first 11 categories are most important as they allow us to distinguish economic from political issues and domestic from European references. Economic issues are covered by the categories welfare, economic liberalism, economic reforms (vague), Euro, anti-bailout and anti-bailout

(conditions). Political issues are covered by the categories democratic renewal, democratic reforms (vague), regionalism, media diversity, and Europe. Note that European issues need to

- 59 - refer explicitly to the European dimension of the question at stake. The coded actors are aggregated according to their party affiliation.

The data analysis proceeds in two steps. At first, we present simple measures of issue salience and polarization. Salience is measured by the share of core sentences related to an issue category in percent of all coded sentences. The indicator for the polarization of party positions is based on a modified version of Taylor and Herman’s index of left-right polarization in the party system, and it ranges from 0 to 1 (for its calculation, see Appendix).

We analyse the two measures because conflicts should be both salient and polarized to structure the dimensions of the political space. Following Kriesi et al. (2008; 2012), we construct the spaces based on the coded issue statements with the help of multi-dimensional scaling (MDS). MDS allows graphically representing the location of parties and issues in a common low-dimensional space. It helps to identify whether and how conflicts over different issues map onto some underlying dimensions (for a more detailed description, see Appendix).

Empirical findings

To assess the importance of conflicts over economic and political issues, we first have a look at the average salience of the principal categories by country. As expected, the results in

Table 2 show that the campaigns are highly focused on economic issues in times of a severe financial and economic crisis. The partisan debate is most concentrated on economics in

Portugal (with more than two-thirds of all statements) and the least in Spain (with around

44%).4 By contrast, Portugal saw the least salient conflicts over political issues. In part, this might be because there was no new party emerging out of the movements in the squares in

4 Analyzing the single campaigns shows that the low salience of economic issues in Spain is due to the 2015 campaign in which political issues trumped over economic ones. In the 2011 cam- paign, by contrast, the focus was as much on economic policies as in Portugal or Greece (see Table 3A of the Online Appendix).

- 60 - Portugal. In the other three countries, conflicts over political issues account for around one- third of all issue statements by parties as reported in the press.

The focus of the agenda on economic matters seems not as surprising during an economic crisis (see also Borghetto and Russo 2018). However, the relatively high salience of political issues supports our first hypothesis: under the impact of contemporaneous crises, the political actors are involved in struggles over both austerity and ways to improve the quality of the democratic process. By contrast, other kinds of substantive policy issues

(ranging from cultural liberalism to immigration and environmental protection) do not get much attention in the mass-mediated partisan debate, which contrasts with previous findings of Kriesi et al. (2008; 2012) for North-western Europe.

Table 3.2 Issue salience by major issue categories

Spain Italy Greece Portugal

Economic 44.1 51.2 60.4 67.5 Domestic (1) 42.7 50.0 30.4 58.9 European (2) 1.4 1.2 30.0 8.6

Political 35.8 34.0 28.1 14.9 Domestic (3) 35.1 30.6 20.8 13.9 European (4) 0.7 3.4 7.3 1.0

Others (5) 20.1 14.8 11.5 17.7

European (2+4) 2.1 4.5 37.3 9.6

N 2142 1521 4002 2947

On the role of Europe in the debates, the findings in Table 3.2 are striking. Apart from

Greece, the campaigns do not constitute critical moments in the politicization of Europe. This result contradicts the expectations of some observers (e.g., Schimmelfennig 2014; Statham and Trenz 2015). Explicit references to European issues amount to less than 10% in

Portugal—with much higher values in 2011 when the country had just accepted its bailout agreement—and even less than 5% in Italy and Spain. In the latter two countries, we rather

- 61 - find constellations of “strategic de-emphasizing” of the European issue in the national electoral arena (Hutter et al. 2016). This mirrors the predominant focus of the challenger parties on the failures of the domestic political class outlined above. For Spain and Portugal, it might also reflect the fact that Europe was traditionally seen as a modernizing and democratizing force (Diez Medrano 2003).

By contrast, and in line with Katsanidou and Otjes’ (2016) case study, references to

European economic (i.e. Eurozone membership, the bailout and its conditions) and political issues were very salient in the partisan debate in Greece. The emphasis on European issues in

Greece mirrors the fact that the country was at the centre of the Euro crisis in many respects.

In fact, Greece was involved in negotiations with its European and international lenders throughout the period covered. The high salience of Europe also concurs with the idea that

Syriza had a clearer dual strategy of criticizing both the failures of the domestic and

European elites. This result does not mean that the struggles over economic and political reforms in the other three countries did not take place in the “shadow” of European constraints. However, the extent to which the European component was explicitly articulated in the electoral arena clearly sets Greece apart from the other three countries.

To examine the structuring capacity of an issue (i.e. the combination of salience and polarization), we now turn to the 18 specific issue categories outlined above. Figure 3.1 shows the two measures by issue and country in a simple scatter plot. The horizontal dimension refers to salience, the vertical dimension to polarization. To ease interpretation, we added lines that indicate the respective means across all issues and countries. The critical issues are the ones in the upper right-hand corner of the figure, that is, the issues that combine an above average salience with an above average polarization. Overall, Figure 3.1 lends additional support to our first hypothesis as conflicts over both economic and political issues are structuring political conflict in Southern Europe. In this respect, it is most significant that

- 62 - the issue of “democratic renewal” (mainly focused on institutional change and on fighting corruption and clientelism) gave rise to highly salient and polarized debates in all countries.

Also, “economic liberalism” is situated in the upper-right corner in Figure 3.1 in all countries except Italy.

Figure 3.1 Issue salience and polarization by specific issue categories

Apart from these similarities, there are marked differences in the type of economic and political issues that were structuring conflicts in the electoral arenas. For Greece, the results again underline the highly contested status of all European issue categories—both economic and political. For Portugal, we observe conflicts over the general need for economic reforms, the bailout (mainly in 2011) and welfare state reforms (primarily in 2015). In Italy, cultural liberalism—in particular, gender equality and gay rights— led to fairly polarized and salient conflicts, too. Finally, Spain stands out for contestation related to regionalism (dominating

- 63 - the 2015 campaign) and the welfare state. We cannot present detailed accounts of the various conflicts in this article. Rather, we want to emphasize here that cultural liberalism in Italy is the only case in which an issue not covered by one of our two broad issue categories— economic or political—was the object of salient and polarized debates.

So far, we have described the most salient and polarized issues in the campaigns, but what we are most interested in is the way the various conflicts relate to each other, that is, how they structure the political space. As explained in the methods’ section, we rely on MDS to represent the distances between parties and issues graphically. Figure 3.2 presents the respective configurations for each country and election separately.

In general, the MDS procedure resulted in two-dimensional solutions for all campaigns covered by the analysis (Figure 3.2). The major dimensions of the partisan spaces in Southern Europe tend to be characterized by amalgamated conflicts over economic and non-economic issues (see also Polk and Rovny 2015; Tsatsanis et al. 2014). More precisely, we find support for our second hypothesis: in all countries except Italy, conflicts over austerity and democratic renewal do not represent independent orthogonal dimensions, but are aligned in the political space. The new challenger parties are closely associated with opposition to both austerity and “old politics” (for an additional systematic analysis, see

Table 4A of the Appendix). This result is illustrated in Figure 3.2 by the fact that the circle that covers the left-wing challengers and the issue democratic renewal is located at the left- wing pole of the economic divide in all countries except for Italy. A key difference is that, in the case of Greece, the main economic conflicts revolve around European measures, but around domestic measures in the other three countries. To substantiate these general remarks, let us take a look at each country in detail.

In Greece, the spaces in all four campaigns are structured around conflicts over domestic and European austerity, as well as over democratic renewal (Figure 2(a)). While the

- 64 - resulting configuration is two dimensional, it is the EU bailout/anti-bailout conflict that dominates the 2012 campaigns and the one in January 2015 as indicated by the spread of all parties along this dimension (see also Katsanidou and Otjes 2016). We have highlighted this main line of conflict by connecting the issues of “anti-bailout” and “(pro-)euro.”5 However, as we want to emphasize here, this conflict overlaps with the divide between the parties that stand for new politics and those that stand for the “old corrupted political class.” As can be seen in Figure 3.2(a), the two established mainstream parties, PASOK and ND, are closely associated with the pro-bailout/anti-democratic renewal pole, while there is a diverse camp of left-wing (The Communist Party KKE and Syriza) and right-wing (ANEL and Golden Dawn) challengers located at the anti-bailout/pro-democratic renewal pole in the May 2012 campaign.

The spatial configuration changed in autumn 2015 when Syriza (now in government) supports the third bailout agreement. For this campaign, Figure 3.2(a) indicates that the main line of conflict is a kind of transformed left-right dimension with Syriza as the main party on the left (close to welfare and democratic renewal) and ND as its main opponent situated close to economic liberalism. Of course, the bailout/anti-bailout division has not disappeared.

However, with the schism within Syriza that led to the formation of the new splinter party

“Popular Unity” in the run-up to the fall 2015 election, the bailout divide no longer overlaps as much with the “new versus old politics” divide.

In Spain, we observe a different configuration which reflects the fact that European issues were hardly debated neither in 2011 nor in 2015 (Figure 3.2(b)). The 2011 campaign is

5 In Greek crisis elections, the classic version of the left–right divide between pro-welfare and economic liberalism is of secondary importance. The secondary status of this divide is due to the rather low salience of welfare-related questions and, most importantly, because the Greek parties tend to agree on a minimum level of welfare protection (in the sense of fighting poverty and supporting social aid) and they all don’t fully embrace the idea of economic liberalism and budgetary rigor at the national level.

- 65 - characterized by a two-dimensional structure, although one dimension dominates the configuration. This divide separates the Socialist Party (PSOE) from the mainstream right

People’s Party (PP). It combines conflicts over welfare and economic liberalism with conflicts over cultural liberalism and defense. The secondary dimension is characterized by questions of regionalism and European integration. As shown in Figure 3.2(b), the radical left

(IU) and the newly created social-liberal Union, Progress and Democracy (UPyD) are closest to democratic renewal in the 2011 election, while the two mainstream parties are equally distant from it. In 2015, we observe even more polarization on the economic left–right divide

(with PP moving closer to economic liberalism and PSOE closer to welfare). This time, the second dimension in Spain is mainly based on regional conflicts in general and Catalonian independence in particular.6

6 While Podemos was in favor of a referendum and more regional autonomy, the other parties—especially Partido Popular— were more opposed to such ideas, as indicated by the fairly large distance of their position to the regional issue. Another minor issue on this secondary dimension was defense, as Ciudadanos and Podemos had different opinions about military interventions debated in reference to the anti-jihadist pact after the Paris attacks a month before the 2015 elections.

- 66 - Figure 3.2 Political spaces in Southern Europe, 2011 to 2015.

- 67 -

Notes: The size of the circles around issues and parties indicates the relative salience of the corresponding actor or object. Selected issue labels: ecolib: economic liberalism; eco-ref: economic reform (vague); demnew: democratic renewal; dem-reform: democratic reform (vague); anti-bail: anti-bailout; anti-bail_c: anti-bailout (conditions) (descriptions of all issue categories can be found in Table 2A of the Online Appendix).

Note that in all graphs only parties with at least 30 observations and issues with at least 3% of all observations were included.

Stress values: (a) Greece. Stress-1: 0.37 (2012 May), 0.29 (2012 June), 0.22 (2015 January), 0.28 (2015 December). (b) Spain. Stress-1: 0.18 (2011), 0.26 (2015). (c) Italy Stress-1: 0.24. (d) Portugal Stress-1: 0.27 (2011), 0.13 (2015).

- 68 - However, most importantly for our argument, there is an even stronger integration of the issue democratic renewal on the left of the economic divide. This becomes obvious when looking at the positioning of the new left-wing challenger Podemos. It is located very close to welfare and democratic renewal as is the old radical left (IU). Also, Figure 3.2(b) indicates that, in line with our third hypothesis, the socialist PSOE in opposition moved to the left and became closely associated with democratic renewal, criticizing corruption and clientelism

(notwithstanding related problems in its own ranks).

The case of Italy in 2013 presents yet another rather exceptional manifestation of how support for democratic renewal is embedded in the political space. More specifically, the

Italian configuration in Figure 3.2(c) does not support our second hypothesis as it comes closest to a two-dimensional space with two orthogonal lines of conflict (see also Conti and

Memoli 2015). As shown before, the main competitors in the party system—ranging from the

Social Democrats (PD) via Berlusconi’s People of Freedom (PDL) to Monti’s Civic Choice

(SC)—are all placed rather closely together in the middle of the economic divide. This reflects the fact that they had all backed at least part of the reform agenda implemented by

Monti’s (incumbent) technocratic government. The results also highlight the fact that M5S is not as closely related to the left as the other new challengers. Rather, as Figure 3.2(c) shows, it is positioned in the centre of the economic divide, far away from economic liberalism and further away from welfare than the two mainstream parties from the left and the right. In the

2013 campaign, its economic program did not attract much attention, and it backed incoherent proposals anyway (Ceccarini and Bordignon 2016: 147 ff.). By contrast, the movement is mainly associated with its opposition to the “casta” and calls for democratic renewal—highlighted by the small size of the circle connecting it with this issue category.

While PD and SC share some of the claims for democratic renewal, the main culprit on this

- 69 - dimension is Berlusconi’s party, PDL, which is located most distant from the issue of democratic renewal.

Finally, for Portugal, Figure 3.2(d) shows different configurations in 2011 and 2015 which has a lot to do with the timing of the bailout in the country and government-opposition dynamics. The Socialists (PS) lost the 2011 election and were succeeded by a mainstream right-wing coalition formed by the Social Democrats (PSD) and the Conservatives (CDS). As indicated in Figure 3.2(d), the structure of the political space is two dimensional for the 2011 campaign. As in Spain and Italy, we observe an economic left–right dimension (indicated by the straight horizontal line). The right, PSD and CD, are situated close to economic liberalism and the radical left, BE and PCP, close to welfare. The governing Socialists are also located at some distance from economic liberalism. Their support for the bailout agreement and attacks from their opponents (blaming them for the political crisis in the country) opened up the second dimension. This second dimension mainly results from a split within the left between the radical left in opposition and the incumbent Socialists. The situation is very different in 2015 when economic issues divide a unified left from a unified right, with positions in favour of institutional change being embedded on the left-hand side of this one dimension. The fact that the PS was by then in the opposition apparently facilitated the unified position of the left regarding both economic and political conflicts (supporting our third hypothesis).7

7 The dominance of such an integrated dimension is in line with findings of Tsatsanis et al. (2014) based on an elite survey during the respective legislative term. The second dimension in the 2015 election campaign is of secondary importance, since the parties do not much address the two issues constituting this dimension (European integration and immigration) and since they are all equally opposed to tougher immigration policies.

- 70 - Conclusions: Restructuring party politics and the rise of a new left?

Based on original media data about national election campaigns, we have assessed the partisan political spaces in four Southern European countries (i.e. Greece, Italy, Portugal, and

Spain) in the Euro crisis. We have developed the argument that, to understand the politics of the crisis, one needs to consider that the countries simultaneously face an economic and a political crisis—a crisis of representation that opposed the old political elite to new challengers. We also argued that both crises have domestic and European components and that given the fact that the challengers forcefully combine opposition to austerity and old politics, the structure of the political spaces is characterized by closely aligned divides. To sum up, we can draw at least four major conclusions from our analysis.

First, putting the accent on similarities, we do find support for the claim that both economic and political issues are key to understanding the structure of political conflict in

Southern Europe in the early 2010s. All countries saw significant combinations of conflicts over austerity and old versus new politics.

Second, apart from Italy, we found a strong alignment of opposition to domestic austerity and calls for democratic renewal in the political spaces. This alignment tends to be stronger if the mainstream left is in opposition. In government, the mainstream left is forced to implement austerity policies, whereas it attempts to adopt economically more left-wing positions and capitalize on the issues of political renewal when in opposition. The Portuguese and Spanish elections in 2015 are telling examples of this mechanism.

Third, the structure of political conflict that is shaping up in all four cases differs from the structure we found in the Northwest of Europe. In North-western Europe, the driving forces of the most recent transformations of party competition has been the new populist right and the conflict structure is characterized by the double impact of the challenges of European integration (perceived as a threat to the sovereign nation state) and immigration (perceived as

- 71 - a threat to the national identity). In the Southern European countries, we have only observed sustained references to the European component of the conflicts in the Greek case. Moreover, the focus on the bailout and its conditions did not politicize Europe in a cultural-identitarian way, but rather replaced the economic left–right divide by a new divide related to European austerity (at least until autumn 2015).

Fourth, there are other major differences in the way the latent potentials induced by the crises have so far been politically articulated in the various countries and in the way they are embedded in pre-existing political divides. Thus, far from claiming that there is a uniform political space in Southern Europe, we have just pointed to the commonalities in this article.

Based on our results, future research should describe and explain in more detail the different manifestations of the general argument. Among others, an interesting avenue for future research is the extremely high salience of EU-related issues in the Greek case. As we speculate in this article, this might be related to the differing strategies of the challengers, the almost constant bargaining process with the European lenders, and ultimately the potential negative diffusion effects that the “Greek case” might have had on electoral campaigns in other countries, such as in Spain in late 2015. However, more research is needed on this front—especially comparative research including countries from North-western and Central-

Eastern Europe as well.

Apart from a more systematic treatment of cross-national differences, another limitation of our study is the focus on “crisis” elections only. Here, follow-up research should extend the time period of the study both back in time and include the elections to come.

Regarding the latter, it may be too early to foretell the outcome of the transformation of the conflict structure in the Southern European party systems. Nonetheless, we would like to end with some speculations. As we interpret the configurations that have been shaping up in the last elections, it is crucial that the main thrust of the new challengers has come from the left.

- 72 - In a way, the crisis has opened up an opportunity for a new left to emerge in Southern

Europe. As our analysis indicates, this new left does not differ much in programmatic terms form the old radical left (the most telling example being the Spanish case). By contrast, it seems that the strong focus on renewing the political system and unfulfilled expectations helped them to gain voters from the old radical left and beyond (e.g. Ramiro and Gomez

2016; Vidal 2018). To the extent that the new challengers from the left dominate the scene, they are the drivers of the integration of the call for democratic renewal into the traditional left–right division. We can see this happening in Greece, where democratic renewal is clearly associated with the left but also in Spain and Portugal, where the fact that the socialists were in opposition in the 2015 has helped to unify the positions of the new leftist challengers with those of the old left. In Italy, finally, the waters have been muddied by the fact that there has been only one crisis election so far. Also, this election has been following up on a technocratic government with which all the major parties were associated. As a result, the political conflict has been the prime mover of the structuration of the party system in the

2013 elections and the conflicts over economic and political issues do not (yet) align. Thus, more evidence from coming electoral campaigns is needed to firmly establish whether and why Italy is indeed a “deviant” case (most likely reasons being the atypical character of the new challenger and the way the economic crisis played out in Italy). However, the other three cases suggest that, ultimately, even this atypical challenger might act as a functional equivalent of the new left and its call for democratic renewal might well align it with the forces on the (centre-)left in the long.

- 73 -

Chapter 4. Out with the Old: Restructuring Spanish Politics

Introduction

The Great Recession in 2008 became a critical turning point for Spanish politics. What used to be an imperfect bi-partisan system with a stable bipolar conflict structure become a fluid landscape with new actors and issues that have rocked the ‘old’ political system. The combination of a political and an economic crisis at the national and European levels profoundly transformed a political system that had been broadly stable. In this article, we focus on the dynamics of transformation of Spanish politics before and after the crises struck.

To do so, we analyse two elections before the beginning of the crisis (2004, 2008) and two elections afterwards (2011, 2015) in depth.

We find that the Spanish story, unlike the Portuguese one (Ferreira and Mendes

2017), but similarly to that of Greece (Altiparmakis 2018), is one of transformation. The elections before the crisis were characterised by two poles identifiable on the left-right scale represented by the two main parties, the Popular Party (PP) and the Socialist Party (PSOE).

Cultural and territorial issues dominated in this period as a satisfactory performance of the economy kept economic issues inconspicuous. After the fall of Lehman Brothers in 2008 and the subsequent European debt crisis in 2009, the public debate shifted dramatically. Initially, the economy was the main concern in the public arena and economic issues dominated the debate in the 2011 election. However, in the 2015 election we observe corruption and political regeneration gaining prominence. The economic crisis coexisted with a political crisis, which was also expressed in a reinvigoration of territorial issues in a new context (i.e.

Catalan independence). In this new multipolar configuration of political conflict, the new

- 75 - radical left (Podemos) drove these transformations by combining advocacy against austerity and for political renewal.

This new scenario revealed different government-opposition dynamics. While in opposition, the mainstream left (PSOE) followed an accommodating strategy and joined the advocacy for political renewal. Moreover, the PP was challenged for the first time by a new centre-right party –Ciudadanos– which combined issues of political renewal, re-centralisation of the territorial organisation and opposition to Catalan separatism. Overall, we find evidence that the political space evolved from a bipolar configuration before the crisis to a multipolar one.

The chapter is structured as follows. First, we contextualise the Spanish cleavage structure and review the main lines of conflict by relying on existing literature. We then proceed to describe the main actors and characteristics of the Spanish party system. Next, the unfolding of the economic and political crisis dynamics is discussed, allowing us to form expectations about the empirical results. These are presented next. A general discussion of the Spanish case concludes the chapter.

Traditional and new divides

After a long period of dictatorship, the societal divisions and the political agency that emerged during the transition to democracy responded to patterns similar to those of the democratic period in the 1930s (Gunther et al. 1986: 14). After some turbulent years marked by a consolidation of the party system and high levels of political violence (Bermeo 1997), the election of 1993 set the lines of conflict for the following decades (Castro 2008; Ruiz

Jimenez 2007). We argue that the Great Recession shook this configuration of the political space, which had been marked by a strong alignment of the dimensions of conflict and a stable party system.

- 76 - Spanish political conflicts have traditionally been articulated along the left-right dimension, which Spanish citizens use for heuristic cues (Torcal and Medina 2007: 277).

This unidimensional meta-structure comprises both economic and cultural dimensions, which are strongly aligned in the case of Spain (Moreno 1999: 29–30; Rovny and Polk 2014: 8;

Vidal 2018) and also encompass traditional cleavages such as religiosity and social class

(Cainzos and Voces 2015; Calvo et al. 2010). On the economic dimension, the poles of the axis are represented by economic liberalism on the right and welfare on the left. At the right- wing pole, the PP stood for the promotion of private initiative and market liberalisation

(Balfour 2005; Michavila Nuñez 2015). At the opposite pole, the socialist party PSOE and the radical left coalition (IU) supported the development of the welfare state and the fight against economic inequality (Mendez 2007; Urquizu 2015). However, the conflictual character of this dimension has at times been reduced due to broad agreements on different social and economic policies such as pensions (Fernández-Albertos and Manzano 2012a).

In contrast to the pacts and agreements reached over welfare policies, cultural policies generally remained controversial, particularly in the pre-crisis period. While the left aligned itself with cultural liberalism, the right embraced patriotism and the defence of traditional values (Ruiz Jimenez 2010, 2007). This was particularly salient during Jose Luis Rodriguez

Zapatero’s first legislature (2004-2008), and led to a peak in polarisation (Balaguer and Sanz

2010). The socialist party pushed for a social rights agenda regarding issues such as equal marriage and abortion rights (Field 2009). Supported by the upper echelons of the Catholic

Church and conservative media outlets, the conservative PP launched a campaign against the government, accusing it of shattering traditional family principles (Sampedro and Seoane

Perez 2009). At the same time, the fight against terrorism became politicised and salient, an issue that linked security concerns with another structuring dimension of Spanish political conflict: the territorial cleavage (Bonet et al. 2010).

- 77 - The territorial cleavage can be considered the third relevant dimension of political conflict necessary to understand Spanish politics (Gunther and Montero 1994; Torcal 2010).

This conflict materialises tensions regarding the territorial organisation of the state, and results in the presence of actors (i.e. regionalist parties) that have small concentrated constituencies and strong veto power in the national institutions. The 1978 constitution introduced a complex four-level territorial organisation1 that attempted to dilute the potential territorial conflict in the national arena and channel it into regional institutions (Colomer

1998; Romero 2012). However, ethno-regionalist parties managed to consolidate their presence in regional institutions and enter the national arena, becoming key actors in government formation.

The territorial dimension has been articulated through a number of issues, which include amongst others: political decentralisation or the distribution of competences between the state and sub-state entities; and interregional solidarity or the distribution of public investment amongst the regions or infrastructure policy.2 Catalonia and the Basque Country stood out in this conflict, claiming self-determination rights using different forms of political action. For instance, one of the salient actors in the Basque self-determination movement was a terrorist organisation (ETA), which led to a close alignment of the territorial conflict with cultural (security) conflicts. The PP campaign transformed the fight against terrorism from a

1 Although the Constitution does not fully set out the nature of the territorial administration, it does mention that besides the national government there will be autonomous communities, provinces and local governments. The inclusion of these figures in the Constitutions is relevant because it crystallises them, given the complexity of the procedure to change the constitution. 2 The territorial dimension of conflict has frequently comprised national infrastructure investment policies, using them as bargaining chips to obtain the support of ethno-regionalist parties to pass other norms. However, this bargaining has also influenced the internal organisation of parties, even the highly hierarchical PP (Castro 2008: 282), as the conflict over hydro-electric resources has frequently shown.

- 78 - valence issue into a positional issue by connecting it to concerns about national identity

(Astudillo 2009; Bonet et al. 2010).

To recapitulate, the structure of political conflict in Spain has three dimensions: economic, cultural and territorial. We have argued that the first two were traditionally aligned along the left-right axis. However, we expect this configuration to have changed with the economic and political crises that unfolded after 2008. However, the political change that

Kriesi et al. (2008, 2012) observed in most of western Europe does not capture the recent political transformations in Spain. ‘New’ cultural issues such as migration, which were at the core of the transformations in NWE, are largely absent from the political agenda in Spain

(Encarnación 2004; Fernández-Albertos and Manzano 2016; Zapata-Barrero 2009). Instead, and in line with the overall argument put forward for southern Europe (Hutter et al. 2018), the combination of deep economic and political crises seems to have triggered changes in the configuration of the party system. Issues concerning political renewal have become more salient, resulting in a reinterpretation of the cultural dimension of conflict. However, before discussing these transformations and the crisis dynamics, in the following section we first review the characteristics of the institutional setting and the party system.

Institutional setting and party system

The 1978 Constitution defined Spain as a parliamentary monarchy with high levels of political decentralisation and a fuzzy delimitation of competences between administrations.

This institutional design also permeated party organisations, translating into different territorial equilibria within the individual parties (Fabre 2010; Hopkin 2009). During the transition to democracy, the territorial organisation was designed to accommodate the expected territorial conflict but also to avoid the tensions that marked the Second Republic and led to the civil war in the 1930s (Gunther et al. 1986). As Linz and Stepan (1996: 99)

- 79 - point out, the problem of stateness posed by Catalan and Basque nationalism was one of the defining challenges of Spain’s transition to democracy, and in their view it was handled comparatively well relative to other countries facing similar situations. Nevertheless, as this chapter will show, the institutional framework set up during the transition to mitigate territorial conflict did not make these tensions disappear, as they later re-surfaced to the forefront of political conflict.

In a gamble similar to the territorial reorganisation, in the 1970s the electoral system was also designed behind a veil of ignorance about the true distribution of prospective voters,3 and it has remained unchanged since (Hopkin 2005: 376–377). Although formally proportional, it includes corrective elements to favour government formation, turning the formally proportional system into an almost majoritarian one (Lago and Montero 2005).

Traditionally, those who benefited most from this electoral system were mainstream office- seekers and ethno-regionalist parties with territorially concentrated support. In contrast, parties such as Izquierda Unida (IU) with dispersed support have found the effective threshold to be much higher, turning the electoral system into a strong obstacle against electoral success (Montero 1998; Urdánoz 2008).

3 The Spanish electoral law devises two parallel electoral systems, one for the Congress and one for the Senate. That of the Congress is the main one. In this election the stakes are higher given that it leads to the appointment of the Prime Minister. Furthermore, its electoral rules are the ones applied by default in regional and local elections whenever the region concerned has not passed its own electoral law, making this combination one of the most frequent in Spain (Montero et al. 1992). Candidates present themselves in closed and blocked lists. The electoral rules provide for 52 constituencies, which fill the 350 seats according to their population size, with a minimum of two seats for the least populated constituencies. Most of the constituencies are allocated between 3 and 5 seats. That is, their size is very small, contributing to the disproportionality. Within constituencies, the seats are allocated using the D’Hondt formula. The Senate is chosen following different rules. Most of the Senate is elected by direct suffrage in provincial constituencies. Voters can tick the individuals for whom they cast their vote, even from different parties (panachage), although this option is not frequently used by voters. The rest of the members of the Senate are elected by the regional parliaments (Villodres and Pereira 2002).

- 80 - The combination of this institutional setting and the political conflict resulted in two main features of the Spanish party system: an imperfect bipartisanship and political parties strongly reliant on sub-national structures. Two main parties dominated the political arena but regional parties also became key players in the case of minority governments (Hopkin 2005).

In terms of organisation, political parties struggled to balance two opposing forces: intra- party discipline and territorial diversity. National leaders were able to exert much control inside parliament, given that Spain is one of the countries with the highest levels of intra- party discipline (Sánchez de Dios 1999: 159). At the same time, subnational entities exerted considerable influence due to their control of resources and regional state institutions and to their contribution to the parties’ electoral resources (Fabre 2010; Pallarés and Keating 2003;

Sánchez de Dios 1999).

Table 4.1 summarises the election results and party system features for the years

2004-2016.4 As the table indicates, the 2011 election shows the first signs of transformation of an otherwise stable party system. With the emergence of the two new challenger parties in

2015 – and the subsequent election in 2016 – we observe an explosion of volatility, a corresponding increase in the effective number of parties and the lowest proportion of votes for the mainstream parties recorded. Although voter turnout declined in the 2011 election, the new parties seemed to drive participation back to pre-crisis levels. All in all, the changes initiated in 2011 seem to have been reinforced in 2015 and consolidated in 2016. The Spanish party system went from being dominated by two main parties to a scenario with four parties, two on each side of the ideological spectrum.

4 The recall election held in 2016 is not included in the analyses of political conflict given that the issues are close to the ones dealt with in the preceding campaign. However, we do include the election in the introduction to describe its outcome and to fully cover the crisis period.

- 81 - Table 4.1 Election results and party system features: percentages and index results.

Election 2004 2008 2011 2015 2016 Election results (vote share)1 Mainstream right 37.71 40.38 45.23 28.89 33.01 PP2 37.71 40.38 45.23 28.89 33.01 Mainstream left 42.59 44.36 29.16 22.14 22.63 PSOE3 42.59 44.36 29.16 22.14 22.63 Mainstream others 4.96 5.01 11.78 3.7 - IU/Unidad Popular (radical left) 4.96 3.81 7.02 3.7 - UPyD (centre-right) - 1.2 4.76 - - Ethnorregionalist4 10.99 7.15 10.42 9.74 9.7 CiU/DyL 3.23 3.06 4.22 2.26 2.01 ERC 2.52 1.14 1.01 2.4 2.66 ICV-EUIA 0.92 - - - - PNV-EAJ 1.65 1.2 1.35 1.2 1.19 EA/Amaiur/Bildu 0.32 - 1.39 0.87 0.77 CC 0.92 0.68 0.59 0.32 0.33 BNG 0.82 0.83 0.76 - - CHA 0.37 - - - - Na-Bai/Gbai 0.24 0.24 0.17 - - Compromís - - 0.52 2.69 2.74 FAC - - 0.41 - - Challenger left 0 0 0 18.12 18.41 Podemos - - - 18.12 18.41 Challenger right 0 0 0 14.04 13.06 Ciudadanos - - - 14.04 13.06 Party system features Turnout (percent) 75.7 73.8 68.9 73.2 66.5 Volatility A 1.3 1.3 2.4 24.8 Volatility B 8.4 2.7 12.2 11.5 Volatility A+B 9.7 4.0 14.6 36.3 Effective number of parties (seats) 2.5 2.3 2.6 4.5 4.2 Mainstream party vote 85.26 89.75 86.17 54.73 55.64 Asymmetry5 9.84 6.59 -13.81 -3.05 -10.38 Polarization (0 to1; own media data) 0.17 0.26 0.19 0.35 - 1 The table only reports the vote share of parties that made it to Parliament. 2 Vote share includes that obtained by parties that have ran in coalition with PP such as UPN, PAR or FAC in 2015. 3 Vote share includes the percentage obtained by PSOE-PSC. Sometimes it is reported separately due to the peculiar relationship between the Catalan party and the national federation. 4 Some of these parties have changed brands across elections. They have been reported together to ease interpretation of the table. 5 This indicator is calculated by subtracting the vote for mainstream right to the mainstream left. Asymmetry = (PSOE+IU/UP) – (PP+UPyD) Sources: Electoral results were obtained from the Parliament's website (http://www.congreso.es/consti/elecciones/generales/index.jsp). Indexes are the result of author's own calculations and those contained in the ParlGov dataset (http://www.parlgov.org/explore/esp/election/).

- 82 - PP and PSOE are the two mainstream parties. The period under study starts when PSOE unexpectedly5 regained office after eight years of PP government. The successes and failures of the PSOE in the 1980s and early 1990s left it a party in deep trouble, from which it only managed to recover in the initial phase of Rodriguez Zapatero’s leadership (Rico 2007;

Urquizu 2015). Zapatero’s first legislature was articulated around the extension of civic rights and a revitalisation of the territorial conflict, while the second legislature focused on the outbreak of the economic crisis.6 PSOE’s perceived incapacity to manage the crisis and be loyal to its constituencies generated tensions that led to severe disputes within the party.

On the right of the spectrum, PP had long been the only contestant. From 2003,

Mariano Rajoy acted as party leader after a personal decision by former president Aznar

(Michavila Nuñez 2015). Rajoy proved to be a political survivor, losing three elections before he won his first in 2011, due to a debacle in the socialist party. During his time in office, he cultivated the image of a trustworthy man (Botti 2013; Martínez and Rodríguez 2015), allowing him to remain in office despite a succession of corruption scandals and the implementation of drastic austerity policies. During his first legislature, some of the regional allies voiced discontent, one of the most notorious being former minister Francisco Alvarez-

Cascos, who founded his own party, FAC. In the 2015 and 2016 elections, however, all of these voices were appeased, and they re-edited existing coalitions such as the one in Navarre with UPN.

Although our results do not show a large presence of regionalist actors in the electoral campaigns,7 their content and development cannot be understood without keeping in mind the

5 This victory was unexpected because the polls predicted that the PP would again win the election, but the 11 March terrorist attacks right before the 2004 election appears to have influenced the electoral result (see Montalvo 2012). 6 See Table 2A in the Appendix for a summary of the governments in the period covered. 7 Regionalist parties are excluded from the empirical analysis due to the low number of observations that are reported in the selected newspapers. Given that the newspapers analysed have a national readership, they tend to over-represent

- 83 - key role of these actors in agenda-setting and government formation (Bonet et al. 2010).

Their territorially concentrated political support made some of them government parties in their regions (mainly PNV and CiU) but vote-seekers in the national arena, able to place their demands on the agenda. The overlap of cleavages reflects the difficulties that national parties face in these regions as they compete on every dimension. In other words, there are leftist ethno-regionalist parties like ERC, Compromís and Bildu, and rightist regionalist parties such as PNV and CiU.

Finally, during the pre-crisis period, two parties became proto-challengers: Izquierda

Unida (radical left) and UPyD (moderate right). Both parties tried to challenge the status quo, advocating for a better quality of democracy, amongst other issues. Izquierda Unida is a coalition of very small radical left parties, including the Communists. The short-lived UPyD was founded by Rosa Díez, a former socialist leader, in 2007 after discontent with the way the party managed issues related to decentralisation. It dissolved after the 2015 election.8

Crises and crisis dynamics

The electoral dynamics of the crisis followed a two-stage process. First, as the economic voting literature would predict, the incumbent was punished and voters turned to the mainstream opposition. PSOE lost the 2011 election to PP, which won an absolute majority

(186 seats and 44.6 percent of the vote). Second, the poor performance of the incumbent and the incapacity of the opposition to offer different policy solutions opened a window of opportunity for new actors to enter the political competition. In the 2015 election, although

national actors at the expense of regional parties. This explains the limited number of cases compared to their prominence in the political arena. 8 UPyD and IU have been labelled ‘proto-challengers, because they intended to play this role in the political system but they were never able to succeed in the way Podemos and Ciudadanos did. UPyD disappeared because Ciudadanos occupied its space, and IU is struggling to find a place, whether as a competitor of Podemos or as an ally.

- 84 - PP remained the most voted party, its vote share decreased substantially (123 seats and 28.7 percent of the vote), while PSOE was unable to capitalise on the shifting vote (90 seats and

22 percent of the vote). Two new forces emerged on either side of the political spectrum –

Podemos (new radical left) and Ciudadanos (centre-right) – which benefitted from these floating voters, putting an end to the political dynamics that had been dominant for decades.

After apparently solid economic growth since the mid-90s, Zapatero’s cabinet faced the financial crisis following the fall of Lehman Brothers in 2008 with optimism, assuring citizens that “the Spanish economy is well prepared to face a situation like the one that we have gone through with the turbulence of the financial sector.”9 However, such optimism proved to be ill-founded (for a review of the causes see Royo 2014). By 2010, the same socialist government had completely shifted its discourse and talked about “special, singular and extraordinary efforts” regarding the austerity policies introduced to face the economic downturn, including raising taxes, cutting salaries, reducing pensions and reigning in social spending.

The implementation of the neoliberal recipes foisted on Spain by the Troika continued with the approval of a labour market reform and a change to the constitution agreed with PP to cap public debt (article 135 of the Constitution). In a fashion very similar to the dynamics behind the collapse of many Latin-American party systems during the 80s as described by

Roberts (2013), the Spanish party system witnessed a neoliberal convergence of the two main parties. PSOE favoured ‘responsibility’ instead of ‘responsiveness’ (Mair 2009), at the cost of being perceived to be an undifferentiated alternative to the mainstream right. In Lupu’s

(2016) words, this produced a brand dilution that reduced the differences between the two mainstream parties.

9 See http://www.elmundo.es/elmundo/2008/09/16/espana/1221587906.html.

- 85 - It was precisely this lack of perceived differences between the two mainstream parties that was voiced with the slogan “PP and PSOE are the same” in the protests that took place in

May 2011, triggered by several organisations, amongst which ‘Democracia Real Ya’ (Real

Democracy Now!) stood out. As in other countries (della Porta 2015), a wave of anti- austerity protests gave way to a new movement: The Indignados or ‘15-M’ movement. Their claims were not limited to anti-austerity but also arose concerns about the Spanish political system and widespread corruption (Anduiza et al. 2014; Calvo 2013). Amidst the turmoil in the protest arena and after a rocky legislature, prime minister Zapatero called an early election for November 2011.

This election was mainly dominated by economic concerns and the government’s lack of ability to restrain unemployment rates. Although the Indignados movement had already raised concerns relating to the political crisis, a soaring destruction of jobs and worsening economic indicators monopolised the political discussion. In fact, unemployment remained the most important problem in the eyes of the citizens.10 At the time of the election, the political crisis did not materialise institutionally, amongst other things because no institutional actor credibly represented preferences for democratic renewal.

PP’s landslide victory, however, neither deterred the introduction of further austerity measures nor appeased a growing sense of political dissatisfaction reflected in the growing levels of mistrust towards key political institutions (García-Albacete et al. 2016a; Muro and

Vidal 2016). In fact, a series of corruption scandals further contributed to the widespread disenchantment. As Orriols and Cordero (2016: 6) point out, the three largest scandals were the Gürtel case, the Bárcenas papers and the ‘black credit cards’ of Caja Madrid/Bankia.

Unsurprisingly, corruption and the political elite soon became the issues of most concern to

10 Corruption only started becoming a main issue in 2011 (see Orriols & Cordero, 2016). Until then, unemployment was the main issue.

- 86 - the Spanish citizenry. Given the new government’s incapacity to ameliorate the economic situation in addition to the corruption scandals, it did not take long for its support to decline.

In 2014, a new party called Podemos ran in the European election and surpassed all expectations in the polls, obtaining close to 8 percent of the vote. With an explicit anti- austerity agenda and an unequivocal populist rhetoric, the new formation presented itself as a party of the people inspired by the Indignados movement. Founded by a group of political science lecturers and leftist activists, Podemos at first tried to appeal to a heterogeneous platform by focusing on political discontent and anti-elitism. However, leftist activists and their own electorate soon pushed them within the radical left, to the despair of those in the party who advocated against left-right politics (Fernández-Albertos 2015b; Galindo et al.

2015). This inconsistency seems to have hindered Podemos’ capacity to fully benefit from the turbulent situation amongst the socialist ranks.

In a very short period the newcomer managed to connect with the wider sense of dissatisfaction deriving from the economic and political crisis, particularly appealing to the young, urban and educated strata of society that showed clear signs of distrust towards the main political institutions (Vidal 2018). In 2015, with the support of smaller factions of the party at the municipal level in large Spanish cities such as Barcelona, Madrid, Valencia and

A Coruña, Podemos obtained 20.7 percent of the vote.

Changes were not limited to the configuration of the left camp in the party system.

The right also saw the rise of a challenger: Ciudadanos, which obtained 14 percent in the

2015 election. Rodríguez Teruel and Barrio (2016) identify two main reasons for this breakthrough. On the one hand, the continued uncovering of corruption scandals within PP facilitated Ciudadanos’ capitalisation on demands for “political renewal, transparency and democratic regeneration” (Rodríguez Teruel and Barrio 2016: 2), a fertile ground also exploited by Podemos. On the other hand, the strengthening of the territorial dimension in

- 87 - Catalonia weakened the Catalan branch of PSOE (PSC), opening a space for an outspoken anti-Catalan-nationalist party.

The Catalan secessionist challenge was, in fact, a fundamental aspect of the political crisis. While the pre-crisis period was dominated by security concerns and disagreement over the way to defeat terrorists in the Basque country, the within-crisis period revolved around

Catalan claims for independence. A reform of the Statute of Autonomy and a ruling by the

Constitutional Court led to a surge in confrontation between the Catalan and national elites.

The arrival of PP in government and the implementation of austerity only added fuel to this conflict (Barrio and Rodríguez-Teruel 2016: 10). A demonstration on 11 September 2013 showed that the conflict had taken on a different meaning amongst citizens, producing a whole new level of support.11 Instead of further autonomy, independence became the main demand. The revival of the Catalan issue has been attributed to a process of ethnic polarisation based on elite-driven motivations (Barrio and Rodríguez-Teruel 2016) in which identity, as well as partisanship, were the main drivers of support for secession (Muñoz and

Tormos 2015). While PP and Ciudadanos had a clear position regarding the territorial organisation, both PSOE and Podemos struggled to place themselves on this issue. Podemos supported holding a referendum but PSOE rejected it, supporting a constitutional reform instead.

The combination of PSOE’s neoliberal turn, the reinvigoration of the Catalan conflict and the emergence of two new parties on different sides of the political spectrum left the socialists in a difficult situation. Despite its initial reluctance to position itself in the political spectrum, Podemos became a solid competitor on the left, while Ciudadanos positioned itself

11 The movement in favour of Catalan independence holds demonstrations every 11 September, the Catalan national day, commemorating the fall of Catalonia to Bourbon troops in 1714. Since 2012, mass demonstrations have been held but none as successful or as numerous as the one held in 2013. That year, the Catalan National Assembly coordinated citizens to make a human chain that stretched over 400 km.

- 88 - on the moderate right. In this new scenario, the socialist party had little choice but to adapt its discourse and distance itself from its recent past, although the new leadership did not necessarily reflect this change.12 If the party was to retain its image of a transformative force in Spanish politics, it had to incorporate the demands for political renewal raised by the new challengers. Being in opposition facilitated PSOE’s adoption of the demands for political renewal.

In sum, the crisis dynamics in Spain were the product of an interplay among a deteriorating economy, a substantial loss of credibility on the part of the traditional parties and a strengthening of the Catalan territorial conflict that followed. Following a period marked by stability, the economic shock introduced these new issues onto the political agenda and brought about volatility and political incertitude. We expect our evidence to show the potential for a new multipolar structure of conflict revolving around austerity, political renewal and the reinvigoration of the territorial conflict. After the socialists’ debacle in 2011, a space opened up for a leftist alternative to tap into feelings of anti-elitism and anti-austerity

(see Hutter et al. 2018). We expect Podemos, the new left challenger, to induce an alignment of left economic issues and democratic renewal. Likewise, a space on the centre-right opened up for a challenger able to combine the transversal claims for democratic renewal (i.e. anti- corruption) with a new strain of the territorial conflict expressed in the Catalan secessionist challenge. Ciudadanos, the new liberal party, seized this new space. Finally, we also expect government-opposition dynamics to be reflected in these transformations.

The content and structure of party competition

In this section, we describe the dynamics of party competition during the electoral campaigns

12 In fact, the new candidate was Alfredo Perez Rubalcaba, who had previously acted as vice-president, minister of the interior and spokesperson for the government, thus not signifying a strong break with past policies.

- 89 - as reported by the media. To observe the impact of the crisis we cover two elections before the outbreak of the Great Recession (2004, 2008) and two elections afterwards (2011,

2015).13 We begin by discussing the salience and politicisation of the main issues that structured the campaigns over time together with the parties’ positioning on these issues. We then explore the transforming configuration of the political spaces.14

A comparison of the salience and politicisation (salience x polarisation) of the main issues covered in the pre-and post-crisis campaigns largely confirms our expectations about the main issues that structured party competition in Spain. Figure 4.1 illustrates the salience and politicisation of a set of aggregate issues during each of the campaigns covered. Figure

4.2 illustrates the party positions with regard to the different issue levels.

Figure 4.1 Salience of issues by election

13 Although some of the results of the 2016 election are also discussed, this election became mostly about parties blaming each other for the need to repeat the election. Moreover, the story told in this chapter is about the transformative nature of the elections in which new challengers emerged, and which the 2016 one did not change. 14 For an overview of the main events during the four campaigns covered, see Table 1A in the Appendix.

- 90 -

In the 2004 campaign, the most salient category is that of ‘old’ cultural issues, which encompasses issues relating to security and defence. One of the key components of this issue is related to terrorism. At the time, the Basque terrorist organisation ETA was being relentlessly pursued by the police with notable detentions. Fighting terrorism thus became the most discussed issue during the campaign. PP tried to build issue ownership on security and the fight against terrorism, particularly after Carod Rovira (the ex-leader of the left-wing

Catalan nationalist party ERC) was found to have met with members of the organisation and received offers of a ceasefire only in Catalonia. Until then, fighting terrorism had been a valence issue but disagreements over the strategy to disband ETA led PP to put the issue on the agenda,15 which explains the low levels of politicisation of the issue reported in Figure

4.1. The participation of Spain in the Iraq war also contributed to the salience of the category, with PSOE promising to withdraw troops and PP planning to keep them in the field. The Iraq

15 PNV/Abertzale left were strongly criticised for their ambivalence regarding terrorism. At this point, the Ajuria Enea Agreement was broken and PP and PSOE stopped being a united front supporting each other’s policies.

- 91 - war issue, however, was secondary to terrorism.16 These issues would linger on into the next legislature, when the socialist government pursued negotiations that led to a permanent ceasefire (Leonisio et al. 2016) and pulled out the troops from Iraq. At the same time, this period was also characterised by growing demands of the Catalan and Basque autonomies for further recognition, which led to a negotiation of new autonomy laws in an attempt to mitigate territorial tensions.

One of the substantial differences between the 2004 and 2008 campaigns is the emphasis on cultural liberalism. During the legislature, the winning party of the 2004 elections, PSOE, focused heavily on civil rights issues such as abortion, allocations for dependents, gender quotas and equal marriage, despite the social resistance rallied by the

Catholic Church. This is reflected in the most salient issues of the 2008 campaign in the

‘new’ cultural category, which encompass all these cultural issues relating to civil rights but also immigration-related issues. The socialist government eased the process of naturalisation and this was not well received by the conservatives. The PP responded by trying to build a position problematising migration, particularly that from Muslim countries. This is clearly reflected in the high degree of polarisation and the correspondingly high level of politicisation of ‘new’ cultural issues in Figure 4.1. Although early signs of the economic crisis already started to show during the 2008 campaign, issues relating to welfare had not yet become polarised. However, the debate around cultural issues also spurred the debate around democratic reform. UPyD ran for the first time in this election to capitalise on this issue,17 although with very modest results, as can be seen from Table 4.1. Overall, these results suggest that the pre-crisis elections were dominated by the territorial and cultural dimensions

16 Note that three days before the election Spain witnessed its worst Al-Qaeda terrorist attack. The proximity to the election day made it largely absent from the sample. 17 Although UPyD ran for the first time in the 2008 election, our results do not include it because of the low number of observations.

- 92 - of conflict, reflecting a bipolar configuration of the political space. Furthermore, the results confirm the highly polarising nature of cultural issues, except for terrorism, which remained salient but not polarised.

In terms of issues, the picture changed radically with the first crisis election in 2011.

Economic liberalism and welfare became the most salient issues by far, completely structuring the conflict in the campaign. The economic crisis was unleashing its full effects and the political debate mostly focused on issues relating to austerity measures and reforms, such as privatising health care, reducing unemployment, estate tax, budgetary rigor, retirement provisions and social justice. The increase in salience of economic issues trumped that of cultural issues. Political renewal, mostly pushed by UPyD, obtained some traction after the 15-M mobilisations, but the political issues were overshadowed by economic concerns.

As anticipated, the 2015 campaign shows a different stage of the crises. Economic liberalism and welfare-related issues were still highly salient. However, the results also show some differences with respect to 2011. For instance, regionalism was no longer related to Basque terrorism but to the Catalan claim for independence. The Catalan secessionist challenge was by far the most recurrent issue during the campaign. Traditionally, territorial organisation had related to democratic and power-sharing views: leftist parties in Spain supported a more decentralising approach while the right preferred more centralisation. Faced with the challenge of a secession referendum, circumstances forced both old and new parties to take positions, as Figure 4.2 shows. Ciudadanos has run on a platform against Catalanism since its foundation and stands strongly against independence. Likewise, PP defended national unity as a strategy to improve its electoral gains outside Catalonia. In the left camp, the limits of power-sharing caused more tensions within the parties. While Podemos advocated holding a referendum but campaigned in favour of remaining, the socialists were divided between

- 93 - different forms of federal solutions.

Figure 4.2 Position of parties by election on main issues

Besides the economy and the Catalan referendum, the political crisis re-emerged, agitated mainly by the new parties. Issues such as political renewal, reforms of the political system, direct democracy and the fight against corruption became an integral part of the political conflict. If in the regionalist conflict Podemos stood alone against all the other national parties, with respect to political renewal it was the PP which was opposed to all the others.

The dominant dynamic was that of government-opposition. The new parties had strong incentives to build platforms on these issues, insisting on their novelty as opposed to existing institutionalised bad practices.

Overall, the crisis campaigns show a picture that is very different from that of the pre- crisis campaigns, with a new radical left party leading these transformations. Cultural issues are replaced by issues relating to political reforms, while regionalist issues change in content and are largely intensified. The 2015 election was dominated by issues concerning political

- 94 - renewal and the Catalan secessionist challenge, confirming the incapacity of the political system to circumscribe the territorial conflict to the subnational arenas. To better observe the conflict alignments, structuring dimensions and government-opposition dynamics, we turn to the more sophisticated methods of multidimensional scaling to observe the relational distances between the different issues and actors across the elections.

- 95 - Political Spaces

Using graphical representations of multidimensional scaling for the issues and party positions

(see Appendix of Chapter 3), we can shed light on the structures of party competition over time and observe how conflicts relate to one another. Figure 4.3 shows the structure of political spaces in the 2004, 2008, 2011 and 2015 campaigns.

Figure 4.3 Political Space in the Spanish campaigns of 2004, 2008, 2011 and 2015 (WMMDS)

- 96 - At first glance, we observe a similar structure for the first two elections. This begins shifting in 2011 and fundamentally transforms in 2015. The evidence shows that in the pre-crisis elections there is a strong integration of the economic and cultural dimensions, the opposite poles being represented by PP –closer to issues such as economic liberalism, defence, security and immigration– and PSOE –closer to welfare, cultural liberalism, education, regionalism and the environment. This largely confirms the expectation of an aligned structure of conflict on the left-right dimension producing a bipolar configuration, which has been reported to be of particular importance in the case of Spain (Rovny and Polk 2014).

A closer look at the configuration in 2011 shows the first signs that a transformation was underway. Both IU and UPyD stand close to political reform and new politics, foreshadowing the structure that ultimately defined the 2015 election. However, as discussed in the previous paragraphs, the 2011 election largely focused on economic issues. It nonetheless represents a transition stage to a period where the mainstream options were not able to truly differentiate from one another and new alternatives emerged.

Lastly, the 2015 campaign data shows the aftermath of a political earthquake. Two new parties enter the scene, raising democratic renewal issues. Three points call for attention.

First, Podemos and Ciudadanos were aligned with issues regarding democratic reform. Even if Ciudadanos successfully capitalised discontent on the right, Podemos, the new left, should be credited with having put the issue on the agenda, institutionalising demands that had frequently been heard at demonstrations since the outbreak of the crisis. Second, the fact that two direct competitors gave salience to issues of democratic renewal forced PSOE, the mainstream left, to reposition itself. It could do this because it was in opposition. Third, the territorial dimension remained a cross-cutting conflict throughout the entire period, even if the content substantively changed. While the new parties shared views on the need for renewal, they were strongly split over how to address the Catalan situation, with Podemos

- 97 - defending the negotiations to hold a referendum and Ciudadanos opposing them on the grounds that consultations are illegal and against national unity. These dynamics created a new configuration of conflict with multiple poles, as can be observed from the enlargement of the grey areas in Figure 4.3. Whereas in the pre-crisis period the main actors aligned along a single dimension, the 2011 and most notably the 2015 elections show how the public debated enlarged around the issues of political renewal and austerity.

Conclusions: Towards a multipolar space?

In line with the ideal-typical scenarios for southern Europe described in Hutter et al. (2018), the story of Spain is that of a process of transformation. The Great Recession triggered a political crisis of major proportions that introduced pronounced changes in the party system configuration and the political agenda. We find that the pre-crisis elections displayed a stable bipolar competition between PP and PSOE, supported by ethno-regionalist parties. Cultural issues, security concerns and the territorial conflict dominated the public debate. In contrast, the crisis elections exhibited volatility and change. This change was driven by demands for political renewal and anti-austerity positions together with the Catalan secessionist crisis facilitating Spain’s transformation to a multipolar party system with the emergence of two new parties: Podemos and Ciudadanos. Overall, the Spanish experience deviates from the general southern European expectations in two main respects. First, a new liberal party emerged linked to the activation of the territorial conflict. Second, no populist radical right party emerged during the crisis period. However, this might be changing with a new party,

Vox, which gained 12 seats in the 2018 regional elections in Andalucía and could make it to the national parliament in the next general elections (see Chapter 7).

As in the other southern European countries, the economy played a prominent role in the first Spanish crisis election (2011). Although some signs of political dissatisfaction and

- 98 - demands for political renewal were already present in the protest arena, the debate centred around austerity measures and reforms. The implementation of austerity policies by the socialist government during the legislature showed a certain degree of convergence between the main office-seeking parties. This lack of a clear political alternative opened the space for new actors to enter the political competition. The second crisis election (2015) reflected this new structural space with the emergence of a new populist left party: Podemos. Comparable to the Greek experience (Altiparmakis 2018), a new party on the left could capitalise on the erosion of the social-democrats due to the implementation of austerity policies enforced by supranational institutions.

While the emergence of Podemos fits the general story of SE quite accurately, the

Spanish case deviates from the expected general pattern as a new centre-liberal party,

Ciudadanos, also emerged. Building on its experience in Catalan politics, it ran as both a response to the Catalan secessionist challenge, which became an integral part of the political crisis, and as another option to renew the political landscape and substitute the existing corrupt elite. A second particularity of the Spanish case is that none of the new challenger parties belonged to the populist radical right. Several interpretations have been advanced for this phenomenon, such as the fact that the territorial dimension in Spain implies that the nativist discourse is already ‘occupied’ (Alonso and Rovira Kaltwasser 2015), that migrants were already amongst the unprotected in a strongly dualised labour market (Fernández-

Albertos and Manzano 2016; Zapata-Barrero 2009) and that there simply was no political agent capable of succeeding (Mendes forthcoming). In this respect, our findings show that despite the PP’s feeble attempt to politicise the immigration issue in the 2008 campaign neither the structural space nor the content of political conflict facilitated any possibility for such a type of party to capitalise on the discontent. Yet, the Catalan secessionist issue (see

- 99 - Chapter 6) might have altered this pattern by opening space for a far-right party to capitalise on Spanish nationalism and against Spanish separatism.

Although the story told in this chapter stops with the emergence of the new challengers in the 2015 election, it is important to add some lines on the events that followed.

The 2015 election resulted in a virtual tie between the four front-runners, to the point that none of them could form a government coalition, and so a new election was held in 2016.

Conflict in this campaign was largely about how the preceding negotiations had developed.

Few substantive policies were discussed. After tense negotiations, PP finally managed to form a government again with the abstention of PSOE and the support of Ciudadanos.

Paradoxically, while the multi-party scenario seems to reflect demands for change in the political system, the ‘out with the old’ discourse driving these transformations did not materialise in the conservatives being kicked out of office, and Mariano Rajoy became president for the third time.

Whether the new actors and lines of conflict will stabilise in the long term remains to be seen. Despite some signs of crystallisation, the demand-side structure of political competition in Spain remains largely unidimensional in terms of economic and cultural issues

(see Chapter 5). Moreover, the issues concerning political renewal, upon which the new parties campaigned, have an expiry date. That is, once parties are established or have assumed government responsibility they can hardly sustain a discourse of political renewal.

While political issues have contributed to a reinterpretation of the cultural dimension, they may in the long-term be replaced once again by ‘new’ cultural issues. Indeed, one possibility is that the current configuration evolves into one like that of the north-western European countries, with a new cultural divide cross-cutting the economic dimension of political competition, encompassing issues such as European integration, nationalism and immigration. This would open up political spaces through a combination of new ‘issue

- 100 - packages’ onto which challengers could consolidate their positions. For the time being, in this rather exceptional transition period we find little evidence, however, of such a convergence to a north-western European type of conflict structure.

- 101 -

- 102 -

Chapter 5. Challenging Business as Usual?

The Rise of New Parties in Spain in Times of Crisis

Introduction

The 2015 Spanish general elections put an end to the two-party system that had existed since the country’s transition to democracy. The two traditional parties that had dominated the

Spanish political arena for over 30 years, the conservative Partido Popular (PP) and the social-democrat Partido Socialista Obrero Español (PSOE), obtained its worst electoral results ever. Together, they amounted to 50.7 percent of the vote, a significant drop since the 2011 elections (73.4 percent), and even more significant since the elections in 2008 (83.8 percent).

The poor results of the two traditional parties were capitalised by two emerging formations: a populist left wing party, Podemos (20.7 percent); and a centre right liberal party, Ciudadanos

(13.9 percent). The drastic political changes that the polls had been foretelling finally consolidated. Why did close to 35% of the electorate chose to vote for new parties in the general elections of 2015? Most importantly, what do the answers to these questions tell us about the magnitude and type of political change that is underway in Spain?

While theories on economic voting predict the punishment of incumbent governments during times of economic crisis, they shed little light as to the political consequences of a wider sense of political dissatisfaction with the overall political system. The mass mobilisations in the protest arena that took place in May 2011 with the ‘Indignados’ movement, and the widespread mistrust towards both national and European political institutions appreciable in public opinion polls, indicate the need to consider alternative explanations beyond reward and punishment models based on economic performance. Using the 2015 and 2016 post-electoral survey data from the CIS to conduct an empirical analysis of

- 103 -

voting behaviour, I argue that the theoretical prism of crisis of representation allows to best understand the emergence of the two new parties in Spain.

The empirical analysis reveals that the economic voting approach to understand the emergence of new parties does not uncover the full story. Although voters of Podemos are more critical of in their perceptions of the economy, neither pocketbook nor sociotropic evaluations seem to clearly distinguish the voters of new parties from voters of the mainstream parties. Instead, it is the attitudes of dissatisfaction towards the political system and the will for political regeneration which make a difference. Moreover, examining the voters of new parties shows that they do not substantially differ in traditional economic and cultural issue positions when compared to the mainstream parties, and that the demand-side structure of political conflict remains largely unidimensional. In this unidimensional structure,

Podemos capitalises the critical voters on the left while Ciudadanos does so on the right.

Lastly, there is an important general divide: it is those concerned with the political situation but especially amongst the young who are more likely to vote for new parties.

While it may be too early to talk about lasting or durable patterns, focusing on Spain as a case study contributes to the understanding of the political consequences of economic shocks, which might allow us to grasp broader political transformations in the shadow of the

Great Recession. The article is structured as follows. The next section contextualises the

Spanish case and reviews theories on the emergence of new parties, economic voting, and crisis of representation; giving way to a set of expectations on the ideological structure and vote characteristics of new parties. The methodology and operationalisation are then briefly discussed before introducing the results. Finally, some concluding remarks are presented.

- 104 -

Party system breakdown and the emergence and success of new parties

The evolution of the Spanish electoral arena during the crisis, as illustrated in Figure 5.1, underwent a two-step process: 1) the punishment of the incumbent in favour of the mainstream opposition; and 2) the emergence of two challenger parties. In the first crisis election (2011) we observe that the mainstream opposition (PP) benefited from the punishment of the incumbent (PSOE), eventually obtaining a clear majority of the votes. In the second crisis elections (2015), however, it was not the mainstream opposition (PSOE) who capitalized on the punishment of the incumbent for the deteriorating economic situation, nor was ultimately the extreme left (IU). Instead, two new parties on the left (Podemos) and the centre (Ciudadanos) suddenly emerged putting a tentative end to the stable Spanish two- party system that was consolidated in the 2016 elections1. The Spanish case perfectly fits the predicament of Hernández and Kriesi (2015), who claim that ‘the cumulated effect of the

Great Recession goes far beyond the short-term punishment of incumbents’, where ‘other mainstream parties which habitually govern hardly benefited from the predicament of the incumbents’ (Hernández and Kriesi 2015: 19).

1 For an analysis on the differences between the 2015 and 2016 elections see Simón (2017).

- 105 -

Figure 5.1 The Emergence of Podemos and Ciudadanos in the Polls

Notes: Dotted vertical lines indicate elections: General Elections 2008; 2011; European Elections 2014 (light gray); General Elections 2015 and 2016. From May 2016, United Left forms a coalition together with Podemos called Unidos Podemos, thus they are represented together. Podemos also includes regional parties such as Marea (Galicia), En Comú Podem (Catalonia) and Compromís (Valencia). Source: CIS Barometer Databank.

How can we explain these unexpected changes in voting behaviour? Most research focusing on the emergence and success of new political parties has relied on either institutional — including the type of electoral system, electoral rules or the structure of the state, amongst others (Harmel and Robertson 1985; Müller-Rommel 1998; Hino 2012; Jackman and Volpert

1996; Abedi 2002; Ladrech 1989)—; sociological explanations —in line with socioeconomic conditions (Dalton et al. 1984; Franklin et al. 1992; Knigge 1998; Lubbers et al. 2002), value orientations of a society (Inglehart 1997; Golder 2003), or cleavage structures (Kriesi et al.

2006, 2008, 2012)—; and spatial strategic models of party competition (Rohrschneider 1993;

Harmel and Svasand 1997; Meguid 2005) following a Downsian perspective whereby the strategic actions of the parties enhance or hamper opportunities for new parties to emerge. For instance, the emergence and success of New Left parties, as well as of the Populist Radical

- 106 -

Right (PRR), has been explained on the basis of socioeconomic explanations such as the

‘enlargement of the tertiary sector, the substantial expansion of educational opportunities during the past two decades, the increasing usage of mass media, and the significant increase in geographical and social mobility of voters’ (Rohrschneider 1993: 683). Other studies have found that economic variables such as unemployment increase the probability of PRR’s success (Knigge 1998; Lubbers et al. 2002).

Most notably, Kriesi et al. (2012, 2008, 2006) have gone a long way in explaining the emergence of a new cultural cleavage (integration-demarcation) that reshapes party systems in North-Western Europe. According to the authors, two waves of transformations rendered a new structure of conflict by which to understand the emergence of the New Left and the PRR:

1) endogenous socio-economic transformations that weakened traditional class and religions cleavages and 2) exogenous processes such as globalisation and EU integration. The story of the new cultural cleavage proposed by Kriesi and his colleagues, however, does not travel well to Southern Europe, where equivalent transformations did not occur. Arguably, the fascist legacies and other supply-side factors in these countries weakened the opportunity structures for the emergence of a new populist right (Alonso and Rovira Kaltwasser 2015), while class struggles sustained the competition between the social-democrats and the communist, thus diminishing the structural opportunities for the New Left. In short, the attenuation of classic cleavages that is appreciable in North-Western Europe (NWE) contrasts with a story of the prevalence of traditional cleavages in the South. In this regard, the structure of conflict would be more similar to Central and Eastern Europe than to NWE

(Rohrschneider and Whitefield 2012).

In the case of Spain, the traditional left-right cleavage has been shown to be particularly strong amongst voters. Rovin and Polik’s (2014) show that there is a high correlation between both socio-cultural and economic issues that can be understood under a single dimension, speaking against the emergence a new cultural divide. Fernández-Albertos

- 107 -

(2017) suggest that this remains the case despite the emergence of new parties. If the structure of conflict remains indeed unidimensional and issue preferences are correlated under a single dimension, we can conclude that the emergence of new parties is not due to new parties offering different configurations of issue ‘packages’ as is the case for the emergence of the

New Left and PRR in North-Western Europe. A first hypothesis can thus be formulated:

Hypothesis 1: The demand-side structure of political conflict in Spain remains unidimensional despite the emergence of new parties (structure of conflict).

Instead, the success of Podemos and Ciudadanos might be better understood by looking at the political consequences of the economic crisis in the short-term. I thus focus on the evolution of the political landscape and the strategic reactions of voters and parties to the economic crisis as an external shock. To that purpose, the following sections review theories of economic voting and crisis of representation that might provide a more suited explanation for the emergence of new parties in Spain.

Economic voting: evaluations of economic performance

Following the classic reward and punishment logic whereby citizens hold their representatives accountable for their actions (Key and Cummings 1966b), literature on economic voting provides some explanations as to the electoral consequences of economic shocks. Incumbent governments, the story goes, will be electorally punished when macroeconomic conditions worsen and rewarded when the economy performs well (Lewis-Beck 1990). To explain variation in the degree of punishment of the incumbent, several conditions have been identified that mediate the strength of the relationship, such as the clarity of responsibility

(Powell and Whitten 1993; Anderson 2000) or the constraints placed on governments insofar they are capable of blame-shifting strategies (Hellwig and Samuels 2007). In Southern

Europe, for instance, Lewis-Beck and Nadeau (2012) show that the effect of economic voting is stronger than in other North-western European countries, arguably due to less complex - 108 -

government coalitions and overall worse performing economies. Although their analysis only covers the period until 2008, more recent research confirms that this trend has continued over the economic crisis. For instance, Hernández and Kriesi (2015) show that economic punishment can be systematically observed throughout Western Europe in the shadow of the

Great Recession.

While macroeconomic studies of economic voting predict strong correlations with the punishment of the incumbent, it is important to look at individual-level economic voting to disentangle the effects on voting behaviour. At the individual level, two types of economic voting can be distinguished. First, voters might evaluate the performance of the economy in general, that is, sociotropic evaluations. These evaluations are, however, often shaped by partisanship. It is expected that supporters of the incumbent government will have better evaluations than those in supporting any other party in opposition (Evans and Pickup 2010; see also Balcells et al. 2015). Second, negative evaluations of the personal economic situation might influence voting behaviour, that is, pocketbook evaluations. A deterioration of the personal economic situation has been shown to have a significant effect for explaining the vote to new challenger parties form both the left and the right (Hobolt and Tilley 2015).

However, as rightly pointed out by Hobolt and Tilley (2015), negative evaluations of the personal economic situation do not shed light in explaining why voters might turn to a challenger party instead of the mainstream opposition. One of the arguments developed in the next section is that both mainstream parties in Spain introduced austerity measures, leading to a situation of brand delusion (Lupu 2016) or neoliberal convergence (Roberts 2017).

Following this argument, evaluations of the personal economic situation should reflect a dissatisfaction with the consensus of mainstream parties in the economic management of the crisis. As such, we would expect that individual-level pocketbook voting reflects a divide between mainstream and new parties, while sociotropic evaluations merely reflect the incumbent-opposition nature of competition. Thus;

- 109 -

Hypothesis 2a: Negative evaluations of the economy in general render a lower probability to vote for the incumbent, but they do not render a higher probability to vote for new parties as opposed to mainstream parties (sociotropic economic voting).

Hypothesis 2b: Negative evaluations of the personal economic finances (pocketbook evaluations) render a higher probability to vote for new parties as opposed to mainstream parties (pocketbook evaluations).

Existing literature on the determinants of vote for Podemos and Ciudadanos in relation to economic voting is somewhat contradictory. On the one hand, Fernández-Albertos (2015a) speculates that Podemos is becoming the party of the economically excluded. This would suggest that pocketbook economic evaluations would be strongly significant in explaining the vote for Podemos. On the other hand, Orriols and Cordero (2016) present evidence showing that pocketbook evaluations do not explain the vote for either Ciudadanos or Podemos, while positive sociotropic evaluations render a higher probability to vote for PP and a negative probability to vote for PSOE. Moreover, Bosch and Durán (2017) show that those more critical with the overall economic situation tend to vote for either Podemos or Ciudadanos, while they only find evidence that those with worse personal economic finances are more likely to vote for Podemos (Bosch and Durán 2017: 7). In any case, economic voting might only be part of the story. In the next section I argue that a crisis of representation unfolded alongside the economic crisis, suggesting that the reasons for the two new parties to emerge might be also found beyond economic-related factors.

The crisis of representation: critical and dissatisfied

Democratic representation consists of a principle-agent relationship between the voters (the principles) and politicians (the agents) which establishes channels for interests to be institutionally represented. Elections are thus the mechanism whereby voters deposit trust in politicians hoping they will defend these interests, which might take different forms, such as delivering material advantages (clientelism) or the pursuit of policy programs (programmatic) - 110 -

(Kitschelt and Wilkinson 2007; Kitschelt 2000). It is precisely through the failure of these linkages between the electors and parties where the crisis of representation occurs and when a window of opportunity opens for new parties (Morgan 2011). From a theoretical standpoint, the failure of linkages between voters and politicians will destabilize party competition. But how are we to recognize a crisis of representation?

Mainwaring (2006) distinguishes two dimensions that constitute a political crisis or crisis of representation: an attitudinal dimension, involving citizens’ perceptions, and a behavioural dimension, which considers ‘actions by citizens rejecting existing mechanisms of democratic representation’ (Mainwaring 2006: 15). The first dimension, attitudinal, also referred to as the latent political potentials (Kriesi 2015: 3), constitutes the degree to which citizens are dissatisfied with the political system and (do not) feel represented by democratic institutions. It is therefore an indicator of the more or less fertility for a political crisis, which only breaks lose when it is manifested in the behavioural dimension, that is, when there is an overt manifestation (Kriesi 2015). The second dimension, behavioural, is concerned with the manifestations in both the formal (electoral volatility, changes in the party system, emergence of new parties) and informal (protests) political arenas. Both dimensions, as explained below, are observable in the case of Spain.

Indicators of political trust towards the key representative national and European political institutions, as well as those of the levels of satisfaction with democracy, are particularly revealing to understand the attitudinal dimension that preceded the emergence of two new parties in Spain. Figure 5.2 illustrates the changes in political and economic indicators in Spain from 2000 to 2015.

- 111 -

Figure 5.2 Political and Economic Crisis Indicators in Spain, 2000-2016

Notes:Vertical lines indicate the beginning of the financial crisis (September, 2008). Sources: All indicators are standarised in percetanges (0-1). Political indicators are obtained from biyearly Eurobarometers (April and October). Situation of the National Economy is a monthly indicator obtaned from CIS (for details on the methodology see http://www.cis.es/cis/opencms/ES/11_barometros/metodologia.html#ISE). Unemployment and youth unemployment is a quarterly measure obtained from Eurostat.

Since the beginning of the economic crisis in September 2008, these indicators followed a remarkable similar pattern (see also Muro and Vidal 2016). The average of satisfaction with the way democracy works in the country decreased substantially after 2008, reaching some of the lowest levels recorded since the transition to democracy. According to Eurobarometer data, while 77% of Spaniards claimed to be very or fairly satisfied with the way democracy works in 2007, only 22% did in 2014. Distrust towards the national parliament and political parties followed an almost identical trend. Although distrust towards political parties had been considerably lower than towards parliament, the levels converged to another historical low in the period after the outbreak of the crisis: less than 10% of the population claimed to trust these institutions. Compared to the levels of trust towards parliament prior to 2008, the decrease is nothing short of spectacular: while close to 50 percent of respondents trusted the

- 112 -

national parliament before the crisis, less than 10 percent of respondents trusted the national parliament at its height2.

While the changes in the political indicators appear to be strongly correlated with changes in the perceptions of the economy (i.e. situation of the national economy) and with objective economic indicators (i.e. unemployment, especially youth unemployment), the economic malaise seemed to trigger a far deeper discontent than with the incumbent’s performance3. Moreover, in the cases of Spain and Portugal, the effect of economic indicators on political mistrust has been found to be rather limited when compared to the effect of the perception of political responsiveness and corruption scandals (Torcal 2014). A simple reward-punishment mechanism based on the performance of the economy alone cannot thus seem to explain the indiscriminate dissatisfaction with the political system that outspread in

Spain during the crisis. Instead, as first pointed out by Krastev (2002) for the Balkans, and later applied to Southern Europe by Bosco and Verney (2012), it is the situation of a

‘democracy without choices’ which produces frustration with the democratic process and aggravates the crisis of political representation. This situation is the outcome of what is best described by the growing tension that governments face between acting ‘responsible’ versus

‘responsive’4 (Mair 2009), or in other words, in dealing with the gap between democracy and efficiency (Scharpf 1999). Under the external constraints of the EMU, the governments of the crisis-ridden South were pressured into adopting a set of fiscal consolidation policies and

2 European political institutions did also not go unpunished and suffered an almost identical drop in trust compared to national institutions, suggesting that the increase in mistrust was not limited by national borders (Muro and Vidal 2016). It is important to note here that domestic issues over reforms of the domestic political system were has much more visibility than issues related to Europe during the post-crisis electoral campaigns (see Hutter et al. 2015). 3 It is important to note, however, that these attitudes do not necessary mean an increase in the illegitimacy towards democracy as a regime, as pointed out by Magalhães (2005) in the case of Portugal. 4 This divide is also brought up in Rohrschneider and Whitefield’s (2012) notion of the strain of representation, which explores the ideological congruence between partisans and independents. The authors conclude that especially in NWE, parties still offer policy alternatives and that they are rooted in their social characteristics.

- 113 -

structural reforms with the objective of reducing debt and regaining access to international credit markets. Incumbents of both sides of the ideological spectrum introduced these unpopular measures to different degrees claiming to act ‘responsibly’ at the cost of being, especially for the left, ‘responsive’ to both their constituencies and ideological principles. The lack of clarity of responsibility for these policies and reforms that stems from the multilevel governing structure of the EU further contributed to a the widespread attribution of blame to all political institutions indistinctively (Hobolt and Tilley 2014).

In the case of Spain, the PSOE announced significant welfare cuts5 and structural reforms as prescribed by the ECB during Zapatero’s government6, such as the labour reform

(Field and Botti 2013), but it also agreed to change the constitution together with PP to cap the budget deficit (Tremlett 2011). Although the retrenchment policies would be much harsher with the government of PP in the following years, the decline in effective political competition between mainstream parties arguably led to a situation of neoliberal convergence similar to that discussed by Roberts (2013) in the case of Latin-America to understand the collapse of the party system (see also Roberts 2017). In short, when major political contenders participated in the process of market liberalization, that is, when centre-left governments either supported or directly introduced reforms in line with the neoliberal orthodoxy, then programmatic dealignment occurred, leading to a situation of neoliberal convergence. The crisis of representation that unfolded in Spain, however, was not limited to the programmatic convergence of the mainstream parties. It was also closely linked to political corruption.

5 For a list of the welfare cuts and reforms in May 2010, see (in Spanish): http://elpais.com/diario/2010/05/13/espana/1273701601_850215.html 6 See letter that the ECB sent to Zapatero in 2011 asking for reforms: https://www.ecb.europa.eu/pub/pdf/other/2011-08-05- letter-from-trichet-and-fernandez-ordonez-to-zapateroen.pdf?e5c1a67f9627c5f087d5c7f02168e0da

- 114 -

The surfacing of corruption scandals became a key issue in the Spanish political agenda in the years following the financial crisis. Concern over corruption and fraud became the second most worrying problem for Spaniards in November 2014 (slightly below unemployment). By then, 63.8 percent of Spanish citizens claimed that corruption and fraud was one of the three most important problems facing the country. Back in March 2008, only an insignificant .2 percent claimed so. Similarly, Spain scored one point lower in the

Corruption Perception Index (CPI) between 2007 and 2016 (from 6.7 to 5.8). Three large corruption scandals linked to the PP (i.e. Gürtel case, the Bárcenas papers and the ‘black credit cards’ of the savings banks Caja Madrid/Bankia) shifted public attention to the issue of corruption (Orriols and Cordero 2016: 6), though there were many others affecting parties of all colours, especially at the regional and local levels (see also Riera et al. 2013: 520).

Political corruption is crucially linked to dissatisfaction with representative institutions. Several studies have found positive correlations between corruptions scandals and political discontent (Ares and Hernández 2017: 2), especially when there are no side benefits to them (Fernández-Vázquez et al. 2016). Moreover, perceptions of corruption are conditional on the economic context (Zechmeister and Zizumbo-Colunga 2013), heightening during times of economic hardship. Other case studies on crisis of representation have shown that, in times of economic crisis, corruption scandals erode voters’ confidence in representative institutions and magnify discontent by highlighting perceived state deficiencies

(Mainwaring 2006). Moreover, recent studies of the Spanish show that being preoccupied with corruption renders a higher probability to vote for Podemos (Bosch and Durán 2017).

This suggests that corruption played an important role in the Spanish crisis of representation and the resulting transformations in the electoral arena, which were first reflected in the protest arena.

Massive mobilisations in the protest arena followed the politics of austerity and corruption scandals. Two small network organizations, Real Democracy Now! (Democracia Real Ya!)

- 115 -

and Youth Without Future (Juventud Sin Futuro), successfully triggered mass demonstrations on May 2011 giving birth to the Indignados protest movement. Neoliberal reforms, welfare cuts and the corrupt political class became the movement’s most visible struggles (della Porta

2015). The participants, which were younger, more educated and less politically involved than in other demonstrations, were also sceptical of traditional political actors such as labour unions and political parties (Anduiza et al. 2014). Despite rejecting the existing political available choices, however, the movement also shared a criticism over political apathy, affecting positively attitudes towards voting (Galais 2014). The so-called ‘Spanish

Revolution’ thus opened a window of opportunity for new formations capable of connecting with the discourse of change that was being voiced in the squares. The message was largely transversal. A survey by Metroscopia7 conducted on July 2011 showed that 81% agreed that the movement had good reasons to protest, while 84% claimed that it raised problems affecting the whole of society. The claims of the protestors, as it seems, trespassed ideological frontiers and largely connected with Spanish society. All of this suggests that the emergence of new parties might be contingent upon the feeling of discontent towards the Spanish political system. Following on from previous literature suggesting that politically dissatisfied citizens tend to have a higher probability to vote for challenger and protest parties (Hobolt and de Vries 2016; Hooghe and Dassonneville 2016), and the discussion on corruption and neoliberal convergence in the Spanish case, we can therefore specify a new hypothesis based on the political crisis:

Hypothesis 3: The higher the voter’s perception of a political crisis, the higher the probability to vote for new parties (political crisis voting).

7 These numbers also include voters of PP. See the survey results here: http://metroscopia.org/wp- content/uploads/2011/06/INFORME-Clima-Social-24ª-Oleada-junio-2011.pdf

- 116 -

Yet, while the drive behind the emergence of these two new parties might be closely related to critical attitudes towards the political system, the question remains as to why two new formations emerged instead of just one. In fact, if we consider the programmatic dealignment for the centre-left as the reasons for the collapse of the party system, we would only expect a party on the left to emerge. Rodríguez Teruel and Barrio (2016) argue that the reasons for

Ciudadanos’ emergence are two-folded. On the one hand, it took a position against predominant nationalism in Catalonia facilitated by the weakening Catalan branch of PSOE

(PSC). On the other hand, it capitalized on ‘unsatisfied demands for political renewal, transparency and democratic regeneration.’ (Rodríguez Teruel and Barrio 2016: 2). Recent research has also pointed to the possibility that stealth democracy attitudes, that is, different preferences on democratic political decision-making procedures, is an important divide between these two parties; Podemos calling for participatory procedures whereas Ciudadanos calling for moderate democratic reforms (Lavezzolo and Ramiro 2017). While the territorial dimension and stealth democracy attitudes might partly answer the question as to the emergence of Ciudadanos, it does not tell us why not just one party was able to capitalise the political dissatisfaction, especially keeping in mind Podemos’ early attempts to shun ideological labels.

One possible answer might be found in the ideological structure of conflict. If the nature of this structure has remained unidimensional as suggested by hypothesis 1, a deeply rooted ideological cleave might have limited the transversality of parties in capitalising political discontent. If so, we might expect that it is those concerned with the political crisis and on the left of the ideological spectrum that vote for Podemos, while it is those most concerned with the political crisis on the right which vote for Ciudadanos. Thus, we can formulate an interaction hypothesis between ideology and critical attitudes towards the political system:

- 117 -

Hypothesis 3a: Those on the left of the ideological spectrum and concerned with the political crisis have a higher probability to vote for Podemos, whereas those on the right and concerned with the political crisis have a higher probability to vote for Ciudadanos (political crisis + ideology).

Beyond the ideological structures, it comes hardly as a surprise that a youth organisation triggered the mobilisations and that the crowd in the 15-M protests was younger than usual.

The management of the economic crisis was particularly harsh amongst the youth, which suffered from particularly high levels of unemployment (up to 55.48% in 2013, see table 2).

The youths’ weak labour position is partly explained by the market segmentation derived from deregulating temporary jobs in presence of strict job security provisions (Noelke 2016:

476). This exceptionally harsh situation for the youth during the crisis, combined with the exposure of corruption scandals affecting the political class, led young Spaniards to ‘echo the crisis to a greater extent than the rest of society’ (García-Albacete et al. 2016b). Thus, a generational divide is expected to be reflected institutionally with the emergence of new parties. This expectation is also substantiated by most research on the determinants of vote for

Podemos and Ciudadanos that have pointed to the existence of a generational divide

(Fernández-Albertos 2015a; Orriols and Cordero 2016; Rodríguez-Teruel et al. 2016).

Bearing this in mind, we can hypothesise that there is an interaction effect between the perceptions of political crisis and age:

Hypothesis 3b: Amongst those concerned with the political crisis, the probability to vote for new parties declines with the increasing age of the voter. (political crisis + age).

- 118 -

Data and methodology

The empirical analysis relies on a pooled dataset of the 2015 and 2016 post-electoral survey study conducted by the Centro de Investigaciones Sociologicas (CIS)8. To test the hypotheses outlined above, multinomial logistic regressions are employed, where the dependent variable is the party voted for in the last elections9: 1) PP, 2) PSOE, 3) Podemos, 4) Ciudadanos, 5)

Other, and 6) Abstention. Table 5.1 summarises the main categories in the dependent variable.

Table 5.1. Parties and Number of Observations in the CIS survey, 2016

Party’s vote share in N party voters Proportion of Party last election (June in sample voters in sample 2016) PP 2438 19.63 33.03 PSOE 2212 17.81 22.66 Podemos* 2083 16.78 21.10 Ciudadanos 1167 9.40 13.05 Total 6984 63.62 89.84

Note: From May 2016, United Left forms a coalition together with Podemos called Unidos Podemos, thus they are represented together. Podemos also includes regional parties such as Marea (Galicia), En Comú Podem (Catalonia) and Compromís (Valencia).

To test the first hypothesis (H1) on the structure of political conflict, I employ a battery of questions included in the surveys on different issues that capture the economic, cultural, and territorial dimensions. These include 1) Public Services vs Tax; 2) Multiculturalism; 3)

Security & Freedom; and 4) Territorial Organisation10. To test for the ideological structure, I first analyse the correlation between the different questions with the left-right scale. Then the average positions of the voters are plotted with fitted lines to observe the correlations between

8 The 2015 post-electoral study 3126 was conducted from January 7th to March 19th 2016 and includes 6.242 interviews. The 2016 post-electoral study 3145 was conducted from July 2nd-21st 2016 and includes 6.175 interviews. Both surveys are representative at the national level stratified and using quotas of gender and age. The interviews were all conducted face-to-face. 9 Last elections refer to the general elections in December 2015 for barometer 3126, and June 2016 for barometer 3145. 10 Details on the operationalisation can be found in Table 2.

- 119 -

parties in the ideological structure.

To test the second hypothesis (H2) on economic voting, two variables are included in the models. First, to capture pocketbook economic voting, a Likert-scale question on the evaluation of the personal economic situation that ranges from ‘Very Good/Good’ to ‘Very

Bad’ is used. Second, to evaluate the effects of sociotropic voting, an index that combines position and salience is crated. The argument is that combining both provides a more accurate reflection of the degree of concern with the economy. Position is determined based on a

Likert-scale question on the general state of the economy that again ranges from ‘Very

Good/Good’ to ‘Very Bad’. The salience variable is constructed hierarchically based on whether the individual claims that the economy or unemployment are the first, second, or third most important problems facing the country. These two variables are combined to produce a normalised ‘Sociotropic’ index.

To test the third hypothesis (H3) on the political crisis, a similar index is constructed combining position and salience. The positional variable is based on a factor that combines the level of satisfaction with democracy with the level of trust towards parliament and political parties11. This is a proxy for being critical with the overall political situation of the country. Once normalised, this variable is then weighted by salience. Similarly, salience is measured hierarchically based on whether the individual claims that corruption is the first, second, or third most important problems facing the country. This renders an index where, one the one extreme, we find those who are most dissatisfied with the political situation and claim that corruption is the most important problem, and on the other extreme, those who fully trust the political system and see corruption as no threat. This same index is also used to

11 A factor analysis of these three indicators shows that all three variables load on a single factor, suggesting that they largely capture the same attitudinal dimension. See Table A.1 in the Appendix for the results of the factor analysis. For robustness purposes, I have also conducted the regressions for each indicator separately weighted by corruption salience (see Table A.2 in the Appendix). Although coefficients vary in the size of the effect, they all report the same direction and level of statistical significance than the overall measure.

- 120 -

test the interaction with ideology (H3a) for which a variable on ideological self-placement ranging from 0 (left) to 10 (right) is used. Similarly, the political crisis index is interacted with age to test for the generational effects (H3b). The models also control for a range of socio-structural and attitudinal variables, including the level of education, gender, region, size of town/city, social class12, region, employment situation, political interest and personal income. A dummy variable is also included to control for the fixed effects of the time when the survey was conducted. Table A.0 in the appendix presents detailed summary statistics of all variables and constructed indices.

Results

To recapitulate, the first hypothesis concerned the structure of political conflict in Spain. The theoretical section discussed the possibility that the emergence of new parties in Spain were possible thanks to the opening of new dimensions of conflict similarly to the emergence of the

New Left and the PRR in NWE (Kriesi et al. 2008, 2012). Against this hypothesis, Rovny and

Polk’s (2014) suggest that there is a unidimensional conflict structure in which the socio- cultural and economic dimensions are strongly correlated in the case of Spain. It could be, however, that this changed with the irruption of Podemos and Ciudadanos. This does not seem to be the case. Figure 5.3 shows how the average positions of the voters for different combinations of policy issues fall under a single dimension.

The demand-side ideological structure appears to be composed of two extremes and two intermediate positions. The extremes are covered by Podemos on the left, combining strong multicultural attitudes with preferences for territorial decentralisation, civil liberties and economic redistribution13; and on the other extreme, combining the exact opposite issue

12 Social class is constructed based on Oesch’s (2006a) class scheme. See Figure A.1 in the Appendix for more details. 13 Note that support for enlarging the welfare state and redistribution are not exactly the same, as pointed out by Fernández- Albertos and Manzano (2012b) in the case of Spain.

- 121 -

preferences, the PP. PSOE and Ciudadanos adopt intermediate positions on the centre-left and centre-right respectively. Conducting a pairwise correlation further suggests that all the issue positions are significantly correlated with the left-right scale (see Table A.3). The results confirm that both economic and socio-cultural issues, particularly the latter, remain strongly correlated with the ideological left-right scale. These results are also in line with the findings of Fernández-Albertos (2017).

Figure 5.3 Average position of voters of main parties in Spain according to various ideological dimensions

Notes: Dotted lines represent the linear fit. All variables have been normalised for simplification purposes.

The second hypothesis (H2) is concerned with economic voting. Two variants were specified.

First, we expected sociotropic economic voting to be driven by partisanship and thus not shed much light on the emergence of new parties (H2a). Second, the evaluation of personal

- 122 -

economic finances was speculated to drive the vote for new parties as a proxy for the rejection of the economic consensus that took place during the management of the economic crisis

(H2b). Table 5.2 shows the marginal effects of three different models based on multinomial logistic regressions for the different parties. Let us for now consider the coefficients on the sociotropic and pocketbook evaluations.

For the sociotropic evaluations, the results are mostly as expected. The probability to declare that the economic situation of the country is doing well significantly increases the probability to vote for the incumbent government, that is, the PP (models 1.A and 3.A). On the contrary, the probability to vote for both PSOE14 (model 1.B) and Podemos (models 1.C and 3.C) increases to a similar degree if the individual claims that the economy is not performing well. All models show that the results are statistically insignificant for

Ciudadanos. With these results we can conclude that, as expected, sociotropic evaluations of the economy are driven by incumbent-opposition dynamics and do not shed much light on the divisions between in the electorate of mainstream and new parties.

For the pocketbook evaluations, the results are mixed. As expected, declaring to have worse personal economic evaluations renders a significantly lower probability to vote for the mainstream parties (models 1.A, 3.A, 1.B and 3.B). Substantively speaking, this is more pronounced in PSOE than in PP (see also model 3 in Tables A.5 and A.6 in the Appendix). As also expected, we find the opposite effect for the Podemos: worse pocketbook evaluations render a higher probability to vote for Podemos (model 3.C, see also Model 3 in Table A.3 in the Appendix)15. However, the results for Ciudadanos defy the expectations as worse

14 Note that the coefficient for sociotropic evaluations becomes insignificant for PSOE once the political crisis index is included. This suggests that once controlling for the effects of the political crisis, sociotropic evaluations become an insignificant predictor. 15 Note that the coefficient for pocketbook evaluations in Table 3, Model 1.C is insignificant. However, in two other models: either controlling for the political crisis (Table 3, Model 3.C), or running a regression with pocketbook evaluations alone (Table A.3, Model 2), the results show statistical significance.

- 123 -

pocketbook evaluations render a negative coefficient (models 1.D and 3.D, see also model 3 in Table A.4 in the Appendix). These results suggest that, while the vote for Podemos is consistent with the expectations for hypothesis H2b, it does not follow for the other new party

Ciudadanos. Therefore, the results only tell part of the story: negative pocketbook evaluations only render a higher probability to vote for Podemos, but not Ciudadanos.

The third hypothesis (H3) concerned the crisis of representation. According to the theory, we could expect that the emergence of new parties answered to a perception of general dissatisfaction with the political system. Let us know focus on the coefficients for the political crisis variable in Table 5.2. In all cases, the results are statistically significant and as expected.

In all models, critical attitudes towards the political system render a higher probability to vote for new parties, whereas they render a lower probability to vote for mainstream parties.

Substantively speaking, Podemos seems to better capture the vote of those dissatisfied with the political situation. This is also confirmed if the different components of the political crisis indicator are run separately in the regressions (see Table A.2 in the Appendix).

If the concern over the political situation is the prime driver of the punishment for old parties, however, why then might we expect the emergence of two new challenger parties and not just one? Given that the ideological structure of political conflict remains unidimensional as tested by the first hypothesis, a variant of the third hypothesis (H3a) suggested that, in this unidimensional structure, voters on the left of the ideological spectrum and dissatisfied with the political system would have a higher probability to vote for Podemos, whereas those on the right for Ciudadanos. The argument here is that the ideological cleavage cuts across the critical attitudes towards the political system.

- 124 -

Figure 5.4 shows the marginal effects of the interaction between the political crisis index and ideology. The results confirm the expectations16: the perception of the political crisis is mediated by ideology to explain the vote for new parties. It is worth noting that

Podemos seems to better attract those most dissatisfied around the centre, though they are unable to reach those disaffected on the right as their strategy of shunning ideological labels was expected to achieve. Given the dissimilar impact that management of the economic crisis had for the different generations in Spain, the final expectation was that there is a significant generational divide that also cuts across perceptions of the political crisis (H3b). Specifically, it was hypothesised that the vote for new parties is driven by those most concerned with the political situation but especially amongst the young.

16 See Model 6 in Tables A.3, A.4, A.5 and A,6 in the Appendix for the marginal effects of the interaction at the mean of ideology.

- 125 -

Table 5.2 Marginal Effects of Multinomial Logistic Regressions on Vote PP PSOE Podemos Ciudadanos Variables 1.A 2.A 3.A 1.B 2.B 3.B 1.C 2.C 3.C 1.D 2.D 3.D

Age 0.002*** 0.002*** 0.002*** 0.004*** 0.003*** 0.004*** -0.003*** -0.003*** -0.003*** -0.002*** -0.002*** -0.002*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Gender -0.007 -0.006 -0.009 0.041*** 0.036*** 0.034*** -0.040*** -0.026*** -0.026*** 0.003 0.002 0.007 (0.008) (0.006) (0.008) (0.010) (0.008) (0.010) (0.009) (0.008) (0.009) (0.008) (0.006) (0.008) Size Town/City -0.005** -0.002 -0.005** -0.004 -0.005* -0.003 0.015*** 0.013*** 0.014*** 0.005** 0.002 0.004* (0.002) (0.002) (0.002) (0.003) (0.003) (0.003) (0.003) (0.002) (0.003) (0.002) (0.002) (0.002) Education (Ref. Higher) Low -0.023 -0.015 -0.012 0.097*** 0.097*** 0.094*** -0.060*** -0.035** -0.032* -0.051*** -0.058*** -0.043*** (0.014) (0.012) (0.014) (0.020) (0.017) (0.020) (0.018) (0.015) (0.018) (0.015) (0.012) (0.015) Middle -0.005 -0.002 0.003 0.049*** 0.047*** 0.047*** -0.020 -0.008 -0.007 -0.027** -0.027*** -0.023** (0.011) (0.009) (0.011) (0.014) (0.012) (0.014) (0.013) (0.011) (0.012) (0.011) (0.010) (0.011) Ideology 0.096*** 0.096*** 0.092*** -0.050*** -0.051*** -0.051*** -0.087*** -0.080*** -0.079*** 0.031*** 0.031*** 0.032*** (0.002) (0.001) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.001) (0.002) Unemployment -0.003 0.002 0.005 0.009 0.006 0.005 0.011 0.012 (0.010) (0.010) (0.013) (0.013) (0.012) (0.012) (0.011) (0.011) Income -0.004* -0.003 -0.001 -0.001 0.003 0.002 0.005** 0.005** (0.002) (0.002) (0.003) (0.003) (0.003) (0.003) (0.002) (0.002) Interest Politics 0.014*** 0.013*** -0.004 -0.009* 0.054*** 0.059*** 0.016*** 0.015*** (0.003) (0.004) (0.004) (0.005) (0.004) (0.004) (0.003) (0.004) Abstention -0.079*** -0.070*** -0.124*** -0.131*** 0.010 0.008 -0.001 0.000 (0.010) (0.011) (0.011) (0.013) (0.012) (0.014) (0.010) (0.012) Economic Voting Sociotropic -0.037** -0.051*** 0.042** 0.027 0.037** 0.066*** -0.006 0.002 (0.015) (0.015) (0.020) (0.020) (0.018) (0.019) (0.016) (0.016) Pocketbook -0.036** -0.027* -0.060*** -0.049*** 0.025 0.034** -0.032** -0.026* (0.015) (0.015) (0.019) (0.019) (0.017) (0.017) (0.015) (0.015) Political Crisis -0.068*** -0.085*** -0.083*** -0.075*** 0.081*** 0.098*** 0.030** 0.029** (0.012) (0.013) (0.015) (0.017) (0.013) (0.016) (0.012) (0.013)

Controls yes yes yes yes yes yes yes yes yes yes yes yes Observations 7,323 9,202 7,280 7,323 9,202 7,280 7,323 9,202 7,280 7,323 9,202 7,280 Notes: Standard errors in parentheses. Levels of statistical significance: *** p<0.01, ** p<0.05, * p<0.

- 126 -

Figure 5.5 shows the marginal effects of the interaction between the political crisis index and age. As expected, the probability to vote for new parties amongst those who are most dissatisfied with the political situation increases as age declines. Again, this effect is substantially stronger in the case of Podemos than in Ciudadanos. Being critical of political system and being younger, on the contrary, renders a drastically lower probability to vote for mainstream parties1. Thus, while age is in itself a significant predictor for the vote towards new parties2, those who are young and critical with the political system are more likely to vote for new parties.

Figure 5.4 Average Marginal Effects of Interaction between Political Crisis Indicator and Ideology on Probability to Vote for PP, PSOE, Podemos, and Ciudadanos

1 See Model 7 in Tables A.3, A.4, A.5 and A,6 in the Appendix for the marginal effects of the interaction at the mean of ideology. 2 Note that all coefficients of age render a statistically significant lower probability to vote for new parties and a higher probability to vote for mainstream parties in all models.

- 127 -

Notes: Based on results from Model 6 in Tables A.3, A.4, A,5 and A,6.

Figure 5.5 Average Marginal Effects of Interaction between Political Crisis Indicator and Age on Probability to Vote for PP, PSOE, Podemos, and Ciudadanos

Notes: Based on results from Model 7 in Tables A.3, A.4, A,5 and A,6 in the Appendix.

Conclusions

The collapse of the party system in Spain observed during the general elections of 2015 and the emergence of two new parties raises important questions that speak directly to the literature on the crisis of representation and cleavage politics. How can we make sense of these transformations and what do they tell us about the ongoing changes in the structure of political conflict in Southern Europe in the shadow of the Great Recession? This article has argued that the key to understand the emergence of both new parties in Spain is through the crisis of representation that unfolded alongside the economic crisis. To get to that conclusion, different hypothesis on the structure of political conflict and economic voting theories were tested using individual-level survey data from the 2015 and 2016 post-electoral survey studies

- 128 -

of CIS.

The finding can be summarised in five points. First, 1) the structure of political conflict in

Spain—for state-wide parties—remains unidimensional despite the emergence of new parties.

This suggest that, unlike the New Left and PRR in North-Western European countries, the emergence of new parties in Spanish is not a consequence of new dimensionalities of conflict.

Second, 2) economic voting approaches only tell part of the story. Sociotropic voting, as expected, only reflected incumbent-opposition dynamics. Pocketbook evaluations, on the other hand, distinguished mainstream parties from Podemos, but not from the other new party

Ciudadanos. It does not seem that evaluations of the personal economic situation are the key feature driving the vote for new parties. Third, 3) dissatisfaction with the overall political system drives the vote for new parties and reduces the probability to vote for mainstream parties. This is consistent with the unfolding of a crisis of representation and its institutional translation. Fourth 4), the unidimensional ideological structure cuts across these critical attitudes insofar those critical and on the left vote for Podemos, whereas those critical on the right vote for Ciudadanos. Fifth and last, 5) there is a strong generational divide that also accounts for the emergence of new parties. It is those critical with the political system, but especially amongst the young who have a higher probability to vote for new parties.

Despite the inherent limitations to case studies, these results constitute a first step to better grasping the political and structural transformations in other countries in the shadow of the Great Recession. The general implication that transcends the Spanish case is that the dimensionality of party competition and socio-demographic divides moderate the behaviour of parties and the response of voters to the economic malaise. Although much remains to be done in comparative perspective, the underlying mechanism presented here is that the impact of the economic crisis—understood as an exogenous shock—will result in political configurations that are mediated by pre-existing demand-side political structures. These political outcomes are, of course, not exempt of other supply-side and contextual factors.

- 129 -

For instance, to stick to two other Southern European examples, the reconfiguration of the Greek party system appears to have been strongly mediated by the issue of the European bailouts and a much more severe economic crisis that tear apart the existing clientelistic linkages (Katsanidou and Otjes 2015b; Hutter et al. 2015). In Portugal, different pathways to democracy (Fernandes 2015), the lack of structural space on the left covered by the communist party PCP, changes in the leadership of the social-democratic party PS, and the economic struggles the country already faced before the economic crisis, are all factors that might have contributed to the seeming stability of the party system in the aftermath of the

Great Recession (Ferreira and Mendes 2017). Future research should further explore the extent to which both supply, demand, and contextual factors explain the different cross- country political outcomes of one of the largest episodes of economic turmoil in modern history.

- 130 -

Chapter 6. Conflict Activation in Times of Crisis: The Case of Catalonia

Introduction

When the first news of an economic crisis landed in Spain by the end of 2008, only 15 percent of Catalans supported the idea of Catalonia becoming an independent state. Six years later, that number escalated to represent half of the Catalan population. Given the entire population of Catalonia (7.5m), this would mean that close to 2.5 million Catalans changed their views with regards to the pillar of a community’s political organization in just over half a decade. With these staggering numbers, it is no surprise that Catalan secessionism became a central concern of Spanish politics, let alone Catalan politics. How could something as fundamental as the preferences for the organization of the State change so rapidly? What led to the activation of this conflict during the aftermath of the Great Recession? What are its political consequences?

To answer these questions, I trace the Catalan process during the period of change and argue that two complementary mechanisms are responsible for the activation of the conflict.

The first has to do with the political strategy of the incumbent government (CiU) led by Artur

Mas that took office in the 2010 regional elections. In order to overcome adverse political circumstances, the incumbent party sought a conflict-displacement strategy by instigating the secessionist conflict. The party was successful in this strategy, brining most of its previous electorate on board and reshaping political conflict as a result. The second mechanism concerns protest voting and political grievances. The profound economic crisis and subsequent crisis of representation at the national level drove many to support independence as means of protest against a discredited political system. With the activation of the conflict, it is shown that Catalan politics underwent a process of ethnic polarisation which is reflected

- 131 -

in the ethnic alignment of the vote starting the in the 2012 regional elections, lingering on to the 2015 and 2017 elections. During this process, it is shown that polarisation in ethnic terms had a strong mobilising effect. This effect was particularly strong for independence supporters at first, but the opposite camp caught up as the conflict consolidated and other parties took an adversarial strategy (i.e. Ciutadans).

The article proceeds as follows. The first section reviews the literature on conflict activation and discusses the strategies that political parties can pursue to influence political conflict. This is followed by an in-depth analysis on the dynamics of conflict activation of

Catalan secessionism focusing on the interaction between the role of the elites, the developments in the public opinion, and the reaction to specific events. The last section consists of an empirical inquiry into the social divides of Catalan society and the activation of the conflict by analysing the evolution of the Catalan electorate. Finally, some concluding remarks are presented.

The Activation of Conflict

What political conflicts structure societies? How and why do these conflicts transform?

These two questions have been one of the primary concerns of comparative politics since, at least, the seminal work of Lipset and Rokkan (1967a). These inquiries respond to the ambition of understanding the formation of cleavage structures, its political translation, and its evolution. Several works have been devoted to this topic in the field of political sociology, a field that Sartori (1969) distinguished from the sociology of politics. The main difference is that, while the former studies the socio-structural architecture of political conflict, the latter deals with the political factors that shape those structures and produce change. Political sociology deals, in essence, with the activation or translation of cleavages into party system structures by focusing on the political factors that produce such transformations.

- 132 -

A notable work in the field of political sociology is that of The Semi-sovereign People, by

Schattschneider (1960), in which he disentangled the political process and proposed two ways in which political conflicts can transform: 1) the development and expansion of conflict and 2) the displacement of subordinate conflicts by dominant conflicts. These two mechanisms point to the idea that societies are confronted not with one, but multiple dimensions of conflict that scuffle with each other to dominate the public agenda (Mair

1997b). This conflict of conflicts, or the “unequal intensity of conflicts”, ultimately determines which issues are politically relevant and, as a consequence, how political systems are structured (Schattschneider 1960: 66).

Yet, not all issues become equally important in defining political systems, nor do all issues become cleavages. As pointed out by Bartolini and Mair (1990), cleavages are composed of three constitutive elements: social structure, values and beliefs, and institutional and organisational agreements. In understanding cleavage formation, structuralist approaches have often thought of this process from a bottom-up perspective, in which the conflict found in the demand-side of politics (i.e. the electoral “mass”) eventually translates into stable cleavages. However, this approach has often neglected the role of political actors (i.e. the supply-side of politics) in influencing the activation of conflict, introducing new issues in the public debate, or even helping to establish new cleavages.

Political actors may seek to influence the political agenda for a number of reasons.

Most understandably, parties and politicians will aim to frame the public debate in a way that is beneficial to their interest. One way to do so is the substitution of conflict, which

Schattschneider argues to be the “most devastating kind of political strategy” (1960: 71). This is because, ultimately, the actor that is capable of introducing a conflict into the political agenda will have an advantage in shaping the political game. The substitution of conflict, or the conflict-displacement strategy, consist of replacing existing issues that dominate the

- 133 -

public debate by other issues in which the party might have issue ownership or a particular advantage. Such a strategy, when successful, can reshape political competition to the advantage of the actor pursuing the strategy and translate into electoral success.

Conflict displacement strategies can take the form of avoidance and instigation.

Avoidance strategies include the undermining of specific issues in an attempt to de- emphasise the line of conflict that is of no advantage to the political actor. Meguid (2005) calls this the dismissive strategy, which refers to a deliberate tactic of “non-action” by a mainstream party in response to an issue introduced by a niche party (2005: 349). This may be combined with strategies that aim to instigate an alternative conflict or change its focus, which may take the form of personalisation strategies —placing the focus on individual political actors as to “distract attention from political issues” (Kriesi et al. 2009: 355); or the adoption of alternative issues.

If, on the one hand, the issue was absent from political competition, then the party may follow an issue-entrepreneurship strategy, defined by Hobolt and De Vries (2015) as “a strategy by which parties mobilize issues that have been largely ignored in party competition and adopt a policy position on the issue that is substantially different from the mainstream status quo” (Hobolt and de Vries 2015: 1161). According to the authors, this strategy is pursued by niche parties and not mainstream ones, since due to the latter’s “overall advantageous position, they have an incentive to reinforce existing patterns of political competition and the policy issues underlying them. Consequently, they are unlikely issue entrepreneurs.” (Hobolt and de Vries 2015: 1161). Moreover, Hobolt and de Vries (2015:

1163) suggest that, since this strategy is a risky one, those parties seeking and issue- entrepreneurial strategy will be political losers, defined by “[lack of] office-holding experience, distance to average party position on the dominant dimension of party competition, and electoral defeat.”.

- 134 -

If, on the other hand, the issue was already present in party competition, we may speak of trespassing. Sides (2006) suggests that parties may trespass by adopting an issue — already owned by another party— when the salience of that issue is high enough in public opinion so that it may prove advantageous. This has been referred to in the literature as riding the wave strategy. The adoption of the issue may be in vague terms or focusing on an attribute of the issue (Kriesi et al. 2009: 355), but the party may also fully engage in that issue and try to steal ownership. Riding the wave differs from issue-entrepreneurship in that the former implies that an issue is already salient in the public opinion, while the latter implies the issue is absent from it. As such, riding the wave is understood as response to a growing demand, while the purpose of issue-entrepreneurship is precisely to create a demand for the new issue in the first place. Table 6.1 summarises the relationship between issues in party competition and the type of displacement strategy.

Table 6.1 Types of displacement strategies by issues in party competition Displacement Strategy Avoidance Instigation New Issue Dismissive Issue entrepreneurship

Existing Issue Dismissive Riding the wave

While riding the wave has been often conceived as a strategy that responds to the changes in the public opinion —i.e. growing salience of a specific issue—, it can also be thought of as a strategy for conflict displacement. Mainstream parties are capable of foreseeing a disadvantageous political context and can opt for trespassing on a specific issue with the hopes to reshape political conflict. This is because the trespassing might create large enough majorities on a specific issue so that the issue itself becomes victorious in the “conflict of conflicts”. If this strategy is successful, then the party may ultimate activate a new line of conflict and secure a dominant place in it.

- 135 -

However, similarly to issue-entrepreneurship, the risks of riding the wave are high.

This is because “the mobilisation [of an alternate issue] may destabilize parties internally, put off certain voters, and jeopardise future coalition negotiations” (Hobolt and de Vries 2015:

1161). Additionally, there is the risk that the party already owning the issue will capitalise on the increasing salience derived from the mainstream’s party trespassing, and the trespasser will not be able to capitalise on the gains. This will depend on the credibility and the reasons given for the trespassing as well as the frame developed by the party to do so. I argue that when the party is at risk of political loss —understood as electoral defeat or the loss of power— then it will be more likely to take on such risks.

There are several contingent factors that might increase the risk of political loss, therefore providing an incentive for a mainstream party to ride the wave. Events such as corruptions scandals or economic crisis —thus being constrained to implement unpopular economic policies— are some examples. These may be combined with contextual factors that provide similar incentives and that might come from outside the party, such as bottom-up pressure from civil society organisations, or within the party, such as internal party factions that might have an ideological incentive to pursue a specific issue. Party militants and civil society organisations close or linked to the party can act as pressuring forces on the party to shift positions on a specific issue, even if the issue in itself is not a key issue for the electorate of the party.

Once a party has changed positions on an issue, there may be important consequences for the electorate. The effect of the partisan elite influencing changes in preferences has been a widely discussed in the literature of the origins of mass opinion and mass-elite linkages with regards to the question “who is cueing whom?” (see Steenbergen et al. 2007; Zaller

1992, 2012). Literature on political psychology explores the mechanisms through which parties might exert such influence (for an extensive review of the literature, see Leeper and

- 136 -

Slothuus 2014). More recent research aims at exploring the causal impact of a party’s change of position of a specific issue on the electorate, confirming that a large share of the electorate does indeed follow the party when it changes its position (Bisgaard and Slothuus 2018).

The case of Catalonia described in the next section provides a real-world example of this complex interaction of factors to explain how the secessionist conflict was activated by analysing the developments in party competition and the actors’ motivations for changing positions and engaging in the issue. This is followed by a section on the consequences of such changes and an empirical analysis of the electorate that, without pretending to engage in the debate over the causality of elite’s influence (given data limitations), shows the evolution of patterns consistent with the above-described capacity of political parties to actively engage in strategies that reshape the political debate and change partisan’s preferences.

The Catalan Revolt: Elites, Context and Events

To understand how the conflict transformed in Catalonia it is imperative to assess the developments of both the demand and the supply side of politics during the time period where change can be observed. In doing so, it is also important to examine the events that unfolded, and the reaction and framings of these events by the different actors. Figure 6.1 illustrates the evolution of Catalans opinion with regards to the relation between Catalonia and Spain. A number of events are also highlighted in the figure, including Spanish general and Catalan parliamentary elections, demonstrations, and other political events (referendums, court’s decisions, etc.) that shaped the public debate in relation to the Spanish-Catalan relations.

The variation in the degree of support for an Independent state observed in Figure 6.1 is substantial. While in the years previous to the economic crisis the number of Catalans that declared to want an independent state amounted to roughly 15 percent, between 2013 and

- 137 -

2015 the corresponding share escalated to almost half of the population (about 45 percent).

As this preference gained weight amongst Catalans, the alternatives of a Federal State or maintaining the status quo (autonomous community) decreased proportionally.

Figure 6.1 Evolution of Catalans opinion with regards to the relation between Spain and Catalonia

Notes: Respondents are asked what they think the relationship between Catalonia and Spain should be. Y-axis indicate the percentage of respondents in each category. Catalan elections are displayed in the bottom of the graph with horizontal labels, whereas general Spanish elections with vertical labels. Events mentioned at the top are explained in the text. Source: Own elaboration from the data of CEO (Centre Estudis d’Opinió). Inspired by a graph published by La Vanguardia (2016).

In particular, a large jump is observed in the year 2012, right between two of the events marked in the figure. These events are the XVI Congress of the right-wing nationalist party

CiU1 (Convergència i Unió) and the 11 of September (national day of Catalonia) which had

1 A federation of two parties: Convergencia Democràtica de Catalunya (CDC), and Unió. After 37 years together, the coalition would finally split three months before the 2015 elections. The party has changed its name several times in the past years.

- 138 -

as a slogan “Catalonia: A New State of Europe”. In order to understand the importance of these two events it is important to take a step back to retrace the evolution of the Catalan political landscape in the years before the crisis.

Several articles have pointed to the sentence of the Statue of Autonomy (SoA) as a defining event for the transformations in the Catalan political sphere (Rico and Liñeira 2014;

Orriols and Rodon 2016; Basta 2017; Muro 2009). Each autonomous community in Spain has its own Statue of Autonomy which consists of a legislative corpus that established the organization of the regional government. In 2003, the tripartite —a coalition of PSC (Partit

Socialista de Catalunya) ERC (Esquerra Republicana de Catalunya) and ICV (Iniciativa-

Verds)2— orchestrated a reform of the SoA that granted more autonomy to the region, which generated a strong opposition from the right. After having been approved in Parliament, the

SOA was ratified in a referendum in 2006, months before the regional elections in November which were characterised by high amount of abstention and dissatisfaction (Lago et al. 2007).

Subsequently, the right-wing Partido Popular (PP) legally contested the reform and presented an appeal to the constitutional court. The final sentence over the SoA would come in June

2010 and declared 14 articles unconstitutional, indicated in Figure 1 as “Resolution CC”. The amendment implied a “decrease [in] powers granted by the approved version.” (Rico and

Liñeira 2014: 261).

The sentencing of the court became a key factor in the change of discourse of the nationalist forces in Catalonia. As argued by Basta (2017: 10), this event was key in redefining the frame of nationalist parties, going from autonomy to independence:

Before the adoption of the 2006 Statute of Autonomy, the dominant Catalan nationalist frame was that of gradual, if contested, advancement of Catalonia’s self-government within the Spanish constitutional framework. Between 2006 and 2010, Catalan

2 Coalition agreement between the nationalist left (ERC), the social democrats (PSC) and the greens/new left (ICV).

- 139 -

secessionists attempted to transform this frame by inducing status-quo nationalist politicians to acknowledge that the gradualist approach to meaningful self-government was at an end. The new diagnostic frame they were proposing was that the Constitutional Court decision on the 2006 Statute, which they predicted would reduce the scope of Catalonia’s autonomy, would decisively demonstrate that Spanish political elites were no longer capable of accommodating Catalan political demands.

The court’s decision provoked a strong reaction in the protest arena. On July 10th, 2010,

Barcelona saw one of its largest demonstration of its democratic history. Under the banner

“we are a nation, we decide”, the constitutional ruling was framed as a violation of the sovereignty of the Catalan people. During this period, we can observe a slight increase in the preferences for an independent state, reaching about 25 percent of the population. After the demonstration, and in a context of “rampant economic recession and growing political discontent” (Rico 2012: 217), president Montilla (PSC) called for snap elections that were held on November 28th, 2010. They resulted in a victory of CiU led by Artur Mas, who would form a minority government. In a sense, these elections constituted the return to normality in Catalan politics, given that, besides the previous two exceptional legislatures governed by the tripartite (2003-2006, led by Pasqual Maragall and 2006-2010, led by José

Montilla) the right-wing party CiU had governed for the previous 23 years. The punishment of the incumbents, as shown by Rico (2012), had much to do with the performance of these parties at the national level, as opposed to the 2006 elections. Table 6.2 shows the results of all Catalan parliamentary elections since 1980s.

The years following the victory of Artur Mas in 2010 until the next snap elections in

November 2012 were politically tumultuous, both in Catalonia and in the rest of Spain, and they are a crucial period for understanding the activation of the secessionist conflict. I put forward two arguments to explain the sudden change of preferences in the territorial organisation during this period. The first is concerned with the change of the centre-right

- 140 -

party CiU. The second is concerned with a growing sense of dissatisfaction with the incumbent government at the national level and the crisis of political representation that unfolded during this period.

Table 6.2 Catalan Parliament Elections Results (percentage over valid votes) 1984–2017

Party 1984 1988 1992 1995 1999 2003 2006 2010 2012 2015 2017 JxCat 21.7 JxSí 39.7 Unió 2.5 CiU 46.8 45.7 46.2 40.9 37.7 30.9 31.5 38.4 30.7 PSC 30.1 29.8 27.5 24.9 37.9 31.2 26.8 18.4 14.4 12.8 13.9 ERC 4.4 4.1 8.0 9.5 8.7 16.4 14.0 7.0 13.7 21.4 PP 7.7 5.3 6.0 13.1 9.5 11.9 10.7 12.4 13.0 8.5 4.2 ICV-EUiA 5.6 7.8 6.5 9.7 2.5 7.3 9.5 7.4 9.9 9.0 7.5 C's 3.0 3.4 7.6 18.0 25.4 CUP 3.5 8.2 4.5 SI 3.3 1.3 CDS 3.8 EUiA 1.4 Others 4.9 2.8 4.7 0.9 1.4 1.4 2.4 6.9 4.5 1.1 1.2 Blank 0.5 0.6 1.2 1.0 0.9 0.9 2.0 2.9 1.5 0.5 0.4 Turnout (%) 64.4 59.4 54.9 63.6 59.2 62.5 56.0 58.8 67.8 75.0 79.1 Note: ICV-EUiA in 2015 runs as Catalunya Sí que Es Pot (CSQEP) and in 2017 as Catalunya en Comú-Podem (CECP). In 2015 CDC (split of CiU) and ERC run together under Junts pel Si (JxSí). Source: idescat. Government of Catalonia

Activation: The Road Towards Independence

Just six months after the regional elections of 2010 in which Artur Mas would become president of the Generalitat, massive protests against austerity policies and in favour of democratic regeneration against the ruling elites flooded Spanish cities, giving birth to the indignados movement. Barcelona was no exception. Indeed, it became one of the main headquarters of the new born movement, which would be the expression of a crisis of political representation linked to the irruption of two new parties at the national arena,

Podemos and Ciudadanos (see Chapter 5), and also to new municipalist movements such as

Barcelona en Comú or En Marea. In Barcelona, the demonstrations included an explicit

- 141 -

rejection of the austerity policies being implemented by the incumbent party CiU. The sudden emergence of the Indignados movement appeared to have temporarily displaced the

‘self-determination’ debate that had been growing since the court ruling over the SOA, introducing a new frame of the ‘the people’ versus ‘the elite’.

The alignment between the anti-elitist discourse and the opposition to austerity made it a particularly dangerous combination for CiU, especially because it represented the political elite that had governed for most of Catalonia’s democratic life and had introduced severe austerity measures during the crisis period. The cuts CiU implemented were substantial and highly unpopular. By May 2012, hardly a year and a half after assuming office, the government of Artur Mas was already announcing the third adjustment plan, this time of

1.500 million, which included the reduction of salary supplements for civil servants, the selling of public assets, increasing university fees, and severe cuts in the health budget3. The commitment to austerity was agreed between both central and regional right-wing governments, an agreement that secured the Catalan government’s access to the regional liquidity fund (FLA).

This adverse political context for CiU worsened due to several corruption scandals affecting the party. The most significant ones were the case Palau, uncovered in 2009, which involved the deviation of 26 million euros of public funds and affected several senior politicians of the party and revealed illegal party financing; the ITV case, uncovered in 2012, regarding the rigging of concessions related to the technical inspection of vehicles for which the secretary general of CDC was condemned for influence peddling; and the public health

3 The cuts in health budget would drop about 1500 million from 2010 to 2014. It is important to note that health was one of the competences of the Generalitat. https://cat.elpais.com/cat/2017/12/10/catalunya/1512922676_645414.html Also, http://www.president.cat/pres_gov/AppJava/president/notespremsa/148663/president-mas-presenta-dajust-obligat-pot- crear-incomoditats-representa-perdua-drets-essencials-serveis-basics.html

- 142 -

scandals, also uncovered in 2012, which included a series of scandals in the public health system concerning years of irregular payments, rigged competitions and political favours that directly involved a MP of the Catalan Parliament of CDC.4

Amidst this adverse political context that aggravated the weight of governing during times of austerity, including several corruption scandals, and the emergence of a new political frame that challenged the elites, CDC held its XIV congress in the small town of Reus on March 23,

2012, as marked in Figure 1. In this congress, titled ‘the national transition’, the party voted by unanimity (99.9 percent) of the votes to start working towards the creation of an independent State. For the first time CDC, which had always embraced moderate Catalanism, would openly support the creation of a new Catalan state. In the words of Oriol Pujol, the secretary general, the process was “irreversible”. The procés (process) towards independence had officially begun with the shift of the mainstream right in favour of secessionism.

This change of position would mark the beginning of a campaign for the regional elections that would take place on the 25th of November 2012, in which CiU adopted a populist rhetoric focusing on “the will of the people” and pursued a strategy of personalisation based on its leader Artur Mas (Barrio et al. 2018). The date of the elections was not decided by chance but strategically placed after September 11, the national day of

Catalonia, where massive demonstrations in favour of independence took place. This time, with CiU openly behind it. Both the timing of the shift in position and the time of the elections suggest that CiU strategically selected the dates to magnify the impact and electorally capitalise its change of position.

4 CDC refers to Convergència Democràtica de Catalunya, which constitutes the majoritarian party in the CiU coalition. Unió, the other partner in the coalition, run separately in the 2015 elections after disputes over support for independence and obtained 2.5% of the vote. Fore a comprehensive list of the corruption scandals affecting CDC see: https://www.casos-aislados.com/Caso-por-partido.php?Partido=CDC

- 143 -

The shift in position of CiU to openly support independence was justified based on the impossibility of further advancing in the ambitions of autonomy of the Catalan people and the negative response from the central government. It is important to note that, contrary to

ERC who had openly supported independence since 1989 (Serrano 2017: 143), CiU had never endorsed it in the past, even if it did not rule out the possibility either. It simply did not consider it “currently relevant” (Newman 1997: 43). In terms of the narrative, Serrano (2017:

148) provides an accurate description of this shift in framing which he calls a “remedial shift towards independence”:

According to the nationalist narrative, the Spanish government’s refusal to negotiate a bilateral process showed the impossibility of finding satisfactory accommodation within Spain, especially after an agenda of recentralization policies on key areas such as education or the failure to make the public investment included in the Statute of 2006. Even though the proposal suggested reaching an agreement with the central government for the referendum to be held under the existing constitutional framework, the situation could eventually legitimate unilateral decisions by the Catalan parliament if the state’s negative position persisted. The official statement of the Parliament underlines the failure of the historical goal of Catalanism, that is, the accommodation of Catalonia within a plurinational Spain, leaving statehood within Europe as the only remaining alternative.

The shift towards openly supporting independence was, therefore, not completely at odds with the party’s past, which had always maintained that possibility. Unlike other cases where a change in position over a certain issue that might be detrimental to a party, this is a case where the narrative was not at odds with the change of position, therefore lowering the risk of being seen as a “copy” to the original party owning the issue.

It is important to mention the weight of civil society organisations in potentially influencing the shift in party position. The pro-independence camp included some highly mobilised civil society organisations such as Omnium cultural (and later ANC — Assamblea

Nacional de Catalunya) that supported several actions such as the not-binding referendums

- 144 -

on independence that 166 cities and towns organised on 13th December 2009 with the participation of 15,000 volunteers (Guibernau 2014: 17; Della Porta et al. 2017). The mobilisation capacity of these organisations made it attractive and safe for a political party to be on their side, and costly to be against them.

Did this change towards supporting an independent state actually respond to a change in the electorate’s preferences? Barrio and Rodríguez-Teruel (2017) tackle this very question by analysing “to what extent the outbidding competition between the main Catalan parties during the 2000s was driven by changes in voters’ attitudes or, instead, by prior elite partisan polarization” (2017: 1777). Using data on party elites and activists from surveys conducted in party conferences, they show that the latter case is true, that is, there was a process of

“outbidding competition without previous significant changes in the opinions and attitudes of the electorate” (2017: 1790). The changes over the territorial preferences illustrated in Figure

1 support this finding, given that the significant change occurs after the XVI conference of

CDC, thus suggesting that the party elites shifted before the electorate. Noteworthily, in the case of national identity in Catalonia, previous research has found elites to be much more radicalised on the issue than the electorate (Baras et al. 2015: 218). In line with this evidence, we therefore expect that the electorate of CiU would have followed the party’s initiative, accounting for much of the change that is observed in Figure 1 after the XVI congress.

Hypothesis 1: Following cues. The electorate of CiU turned more pro-independence following the party’s change in position

So far, I have argued that the adverse political context for CDC —such as the indignados mobilisations, the weight of governing during the crisis, or corruptions scandals—, together with bottom-up pressure from civil society organisations and a more nationalistic elite, provided incentives for CDC to change its position on the issue of secessionism. This change can be considered as trespassing, since the mainstream centre-left nationalist competitor ERC

- 145 -

had advocated secessionism for a long time. However, the change of position of this party might not reveal the full story. In fact, there is a second mechanism at play that might also explain the growing vote in favour of independence: protest voting.

The rising support for independence in Catalonia cannot be fully understood without the developments of Spanish politics in the aftermath of the crisis. The economic crisis gave way to a political crisis throughout the Spanish territory, resulting in high levels of distrust towards political institutions. The perception that the left-wing Zapatero government (2008-

2011) did not introduce any different economic policies than its main right-wing competitor would have introduced, generated a crisis of representation that would be voiced by the 15-M as “PP and PSOE are the same”. Yet, even after the widespread 2011 protest, the right-wing

PP won the Spanish elections in November. The victory of PP constituted a new opportunity for the pro-secessionist discourse, especially since nationalism in Catalonia is not orthogonal to the left-right dimension but aligned with the left (Dinas 2012) due the historical association between Spanish nationalism and the Francoist regime (Muro 2018).

The promise of a different political system proposed by Catalan secessionism that contested, to the eyes of many, an irreformable Spanish state, constituted an attractive political project for those most dissatisfied with the functioning of democracy. I thus expect political grievances arising from the crisis of representation (see Chapter 5) to be correlated with support for independence. I expect this correlation, however, to be more pronounced amongst those with instrumental reasons for independence as opposed to identitarian reasons.

Since it has been repeatedly shown that support for independence is strongly correlated with national identity (as explained below), it is expected that the protest mechanism will be more pronounced amongst those with dual identities, which is taken here as a proxy for having instrumental reasons for wanting independence. Thus;

- 146 -

Hypothesis 2: Protest Voting. Support for independence is mediated by political grievances, especially amongst those with dual identities.

In sum, it was argued that two mechanisms produced the large shift in the electorate and activated the secessionist conflict. The first is the change of position of CiU, which is explained by the adverse circumstances of the party at the time and the need to reshape political conflict to sustain power, and which would account for the large shift in preferences towards secession from its electorate. Second, the political grievances stemming from the crisis of representation also drove many to support the creation of a new Catalan state. What were the consequences of such activation?

Consequences: Ethnic alignment and polarisation

The activation of the secessionist conflict in Catalonia produced profound changes in the structure of conflict, which are expected to have translated into the electorate. In this new scenario of conflict, the dominant political divide would no longer be the left-right divide but the divide about secessionism. Because support for secessionism has been shown to be strongly correlated with national identity (Muñoz and Tormos 2015; Orriols and Rodon

2016), the activation of the secessionist conflict is expected to have activated a process of ethnic polarisation amongst the electorate. Comparing the 2010 and 2012 at the aggregate level, Serrano and Bonillo (2017) show the “existence of an ethnic alignment of vote and that this relation grows as reflected in the variation between the elections of 2010 and 2012”

(2017: 385). Considering that the conflict has by no means decayed in intensity in later elections (2015 and 2017), we would also expect this ethnic polarisation to have become stronger in the electorate at the individual level. In other words, we would expect that national identity would become a stronger predictor of voting behaviour.

Hypothesis 3: Ethnic polarisation. National identity becomes a more relevant factor in predicting the vote from the 2012 elections onwards.

- 147 -

With the polarisation of the conflict, we could also expect a larger turnout in the elections from both ‘identity’ camps. This is because, just as decline in polarisation of conflict has a negative effect on turnout (Steiner and Martin 2012), increasing polarisation may, on the contrary, lead to higher turnout, especially when something as fundamental as the creation of a new state is at stake (Andersen et al. 2014). I expect that the pro-secessionist camp mobilised first, and that the anti-secessionist camp caught up only after some time. This is the case for two reasons. First, because the Spanish identity in Catalonia may be correlated with lower education levels, which known to correlate with lower political participation and turnout (Persson 2015). Second, because the activation of the secessionist conflict is likely to have triggered an adversarial strategy from the opposition, namely Ciudadanos, who would have campaigned strongly against secessionism, thus attracting the vote of those strongly opposed to it as the issue became more salient.

Hypothesis 4: Ethnic mobilisation. Polarisation leads to higher turnout in a sequential manner, first by the pro-secessionist electorate, then followed by those against secession.

Data and Methods

This study relies on three different sources of survey data that include different elements that are of interest to the analysis, namely the CEO (Centre d’Estudis d’opinió) from 2006-2018, which consists of a periodical survey on electoral behaviour conducted by the Catalan regional Government; the post-electoral CIS (Centro de Investigaciones Sociológicas) surveys from 2006-2017, which is the Spanish state-wide equivalent; and the ICPS (Institut de Ciències Socials i Polítiques) surveys available from 1991-2016, an institute affiliated with the Autonomous University of Barcelona. Each of the surveys provides advantages to tackle different aspects. The CEO survey is conducted up to three times a year and it allows for a more nuanced view of the changes in support for key questions regarding secessionism.

- 148 -

However, the methodology of the survey changed from phone interviews to face-to-face interviews in 2013 and as such there is a break in the series that limits temporal analysis. The

ICPS survey is only conducted once a year but it has an extensive questionnaire and more specific questions than the other two surveys with regards to Catalan’s opinion and preferences over independence. It also has the advantage of going back until the early 90s.

Finally, the CIS provides post-electoral surveys close to the election dates with a large sample but with a limited number of questions.

Ideally this study would have benefited from individual-level panel data to shed light on the causal mechanisms behind the activation of the conflict, but to my knowledge such data does not exist. Each of the hypotheses are tested using the survey data that best allowed to isolate the effect. Replication amongst datasets was conducted when possible (results available upon request). Regression analysis is based on logistic regressions and includes the most used socio-economic control variables, namely gender, age, education, province and size of municipality, as well as ideological variables when pertinent such as the left-right scale or identity variables such as national identity.

National identity: The Linz-Moreno Question

The question on national identity is based on the Linz-Moreno question, which was first conceptualised by Linz et al., (1973) for the case of Spain. This question includes five categories, asking respondents to choose amongst a range of categories between “I feel only

[national]” and “I feel only [regional]”, including a dual identity category in which respondents report being equally national and regional. This question has been used in several studies of nationalism and has been recently scrutinized in depth by Guinjoan and

Rodon (2016). The categories of “I feel only Spanish” and “I feel more Spanish than

Catalan” have been lumped together following other studies (Muñoz and Tormos 2015; Rico

- 149 -

and Jennings 2012; Muro and Vlaskamp 2016). This is because there are often very few observations in those categories and because there are no theoretically different expectations from those respondents.

One of the problems of using the Linz-Moreno question to observe changes in the support for independence is that identity might not be exogenous to the preference for territorial organisation. In other words, supporting independence might also lead an individual to declare itself ‘Only Catalan’. For this reason, I have replicated whenever possible the results with two other variables that serve as proxies for identity, namely the language used at home and the place of origin of parents. These proxies for identity should be more stable over time and reduce the endogeneity of the relationship.

Empirical Analysis: Activation and Consequences of Secessionism in Catalonia

Following cues

The first hypothesis suggested that CiU’s electorate turned more pro-independence with the party’s change of position. I first analyse the share of pro-independence supporters over time.

Comparing support for independence across waves amongst those who claim to have voted for a party in the previous elections allows us to approximate the degree to which the electorate of each party changed preferences over the issue of secessionism. Figure 6.2 shows the percentage of those favouring independence by the party voted in the last Catalan parliamentary elections.

- 150 -

Figure 6.2 Share of pro-independence citizens by party voted in last Catalan elections, 2006- 2014

Notes: Own elaboration from CEO Barometer (2018). Each dot for line “CiU” represents an individual survey study. Vertical lines represent Catalan Parliamentary elections. Source: CEO, 2006-2013.

There are at least three aspects worth highlighting. The first is the constant increase of ERC’s electorate in preferring an independent state. Although ERC always advocated for an independent state, only 40 percent of its electorate supported independence before the Great

Recession. This number escalated close to 90 percent in 2013. Second, the constituency of

ICV, a left-federalist party, underwent an important shift prior to the 2012 elections. The percentage of individuals that reported having voted for ICV in 2010 and that supported independence increased to 45 percent in October 2012 from barely 25 percent in June of the same year. However, this number is reduced down to 30 percent after the 2012 elections.

Third, the percentage of pro-independence supporters amongst those who claim to have voted

- 151 -

for CiU in the 2010 elections, increased from 28 percent in January 2011 up to 60 percent in

October 2012, right before the November 2012 elections. The largest change occurs between

June 2012 and October 2012, where the percentage increases by nearly 20 points. If we consider the period between the 2010 and 2012, the largest shift comes from the electorate of

CiU.

These results suggest that a substantial share of the electorate of CiU in 2010 changed their preferences with regards to the organisation of the State alongside the party’s change in position, shortly after the 2012 elections. Modelling support of independence provides the same results, insofar as the marginal effect of having voted for CiU in the 2010 elections becomes a significant predictor of supporting independence from October 2012 onwards

(Figure A1 in the appendix). Moreover, building on V.O. Key’s (1966b) distinction between types of voters, we can identify some of the elements that distinguish the voters that voted in

2010 and again in 2012 for CiU —the loyalists (coded as 1)— versus those who changed party in 2012 —the leavers (coded as 0). If support for independence is a positive and significant predictor of loyalty, we can then conclude that CiU appealed most to the electorate that was either already in favour of secessionism or that became secessionist with the party’s change of position. Table 6.3 shows the results of various model specifications and confirms that those more in favour of secessionism decided to vote again in 2012 for

CiU.

The results suggest that that those who support independence are between 12 to 20 percentage points (models 1-3, Table 6.3) more likely to have remained loyal to CiU in the

2012 elections5. These results are robust through a number of different model specifications

5 In terms of odds ratios, the results indicate that those who support independence are 3 to 4 times more likely to have remained loyal to CiU (as opposed to having switched) than those who do not (see table A2, models 2-4 in the appendix).

- 152 -

and controls, including leadership evaluation, party identification (PID), salience of the territorial reform, and national self-identification. Replicating similar models for other available survey data (CIS) provides similar results6. Given that the party only lost 7-8 percent of the vote between 2010 and 2012, it is safe to conclude that a large share changed their preferences with regards to secession. Moreover, is it also possible to see the extent to which those who had not voted for CiU and did in 2012 —the newcomers— compare to those who did not vote for CiU neither in 2010 or 2012 —the opposers. The newcomers have a 5 to 9 percent higher probability of being pro-independence, depending on the model specification (models 4-6, Table 6.3), suggesting that the party also managed to attract new voters on the basis of its change towards secessionism, mostly from its competitor ERC.

So far, I have shown that CiU’s change of position regarding secessionism was accompanied by a change in its electorate towards more pro-secessionist positions, and that those who stayed loyal to the party in 2012 were those most in favour of independence. It is important to note, however, that in order to establish the causality from the party’s change in position to the change in the preference, more sophisticated panel data would be required (not available to my knowledge). There is some additional evidence that supports this interpretation. First, Barrio and Rodríguez-Teruel (2016) show that the ethnic polarisation is likely to be an elite driven process by looking at national identification data from politicians and activist. Second, it is important to note that CiU held its XVI congress in March 2012, thus preceding the large shift in preferences observed in the following months, and that coincided with the September 2012 pro-independence demonstration.

6 Results available upon request.

- 153 -

Table 6.3 Marginal effects of leavers versus loyalists and newcomers vs. opposers of CiU, 2010-2012

Loyalists vs Leavers Newcomers vs. Opposers Variables (1) (2) (3) (4) (5) (6)

Pro-Independence 0.197** 0.176*** 0.127** 0.086*** 0.044 0.054* (0.082) (0.064) (0.065) (0.032) (0.028) (0.031)

Salience 0.076** 0.051 0.024 0.022 0.014 0.011 (0.034) (0.034) (0.039) (0.017) (0.015) (0.017) National Sentiment (ref: Only/More Spanish) Dual 0.175 0.294 Ref. Ref. (0.228) (0.228) More Catalan 0.365 0.420* 0.099*** 0.074* (0.246) (0.240) (0.036) (0.038) Only Catalan 0.369 0.402* 0.023 -0.023 (0.248) (0.243) (0.034) (0.036)

Evaluation A. Mas 0.013 0.011 0.011** 0.013** (0.012) (0.012) (0.005) (0.006)

PID 0.138*** 0.142*** 0.050*** 0.053*** (0.034) (0.036) (0.012) (0.012)

Controls ✔ ✔ ✔ ✔ ✔ ✔

Observations 221 222 221 549 618 521 Notes: For models 1-3 the dependent variable is loyalists (1) vs. leavers (0). For models 4-6 the dependent variable is newcomers (1) vs. opposers (0). Standard errors in parentheses (*** p<0.01, ** p<0.05, * p<0.1). Controls include: gender, age, education, size of the municipality, province, left-right scale. In models 5-8 there are no observations for ‘Only/More Spanish’ in ‘newcomers’. As such, the reference category becomes ‘dual’. Source: ICPS, 2012.

Protest Voting

Although most of the change came from the electorate of CiU, it was also hypothesised that the growth of support for independence was linked to the political grievances that aggravated in the aftermath of the economic crisis (H2). The change towards supporting independence could thus be associated with a second mechanism beyond following elite cues of CiU in activating ethnic-polarisation, namely a growing feeling of detachment from the central government, which substantially increased in average over time, especially from 2011 to

2012 (see figure A4 in the appendix). In short, supporting independence could be a way of

- 154 -

voicing protest7. To test this argument, I check whether it is particularly among those with dual identities that political grievances drive the desire for an independent state.

Figure 6.3a shows the predicted probabilities of support for independence over different levels of satisfaction with democracy. While the preference for an independent state is strongly correlated with national identity, especially for those categories of ‘only Catalan’ and ‘more Catalan than Spanish’, the mediating effect of political grievances in predicting support for independence varies significantly. More specifically, those who identify as ‘only

Catalan’ only show a slightly higher probability to support an independent state as they are more dissatisfied with the functioning of democracy. By contrast, the slope is much steeper for those who identify as ‘more Catalan than Spanish’, increasing twenty percentage points from 30% to 50% as the self-reported satisfaction with democracy decreases. For voters with a Spanish identity or who equally identify as Spaniard or Catalan, political dissatisfaction does not make a difference at all.

As pointed out by Guinjoan and Rodon (2016), the ‘dual’ category, ‘as Spanish as

Catalan’, is associated with pro-Spanish identity positions, whereas ‘more Catalan than

Spanish’ captures to a greater extent the duality due to historical associations with Spanish nationalism (Dinas 2012), which explains why the ‘dual’ category is associated with no support for secessionism together with the ‘only Spanish’ category. The results hold after controlling for several political (ideology, party identification) and identity-based variables

(national identity) across a number of model specifications and using different data sources8

(see table A3 in the appendix). Although we cannot be certain of the causal effect —i.e. whether supporting independence leads to lower satisfaction with democracy or the other

7 Note that there are a non-negligible number of citizens who would vote for independence even if their preferred model of territorial organisation is a federal state, although this number declines over time (see figure A.2 in the appendix). 8 Replicating the model for CEO and CIS provides similar results with regards to central government evaluation. Data available upon request.

- 155 -

way around—, the correlation suggests that there are strong protest dynamics at play in the

changing support towards secessionism of this specific category of voters.

Figures 6.3 Predicted probabilities of vote for independence by preference for the territorial organisation by political grievances

Notes: Based on Model 4 in Table A3 in the appendix. Source: CEO, 2010-2013

Towards Ethnic Alignment

The crisis of representation and the change in position of CiU might explain the shift in

preferences over independence, but what exactly triggered this change? Hypothesis 3

suggested that ethnic polarisation became an increasingly important factor in predicting

voting behaviour in Catalonia since the beginning of the procés (2012 –). Serrano and

Bonillo (2017) show this to be the case at the aggregate level. In the context of growing

political polarisation, they find evidence of ethnic voting alignment from the 2010 to the

- 156 -

2012 election. However, there are problems with using aggregate data, namely the danger of ecological fallacies (Przeworski and Teune 1970). As Serrano and Bonillo point out, they do not “aim at inferring individual behaviour as much as detecting patterns of change at the aggregate level”. I test this hypothesis at the individual. Using post electoral data, Figure 6.4 shows the predicted probabilities of national identity on voting for pro-secessionist parties

(CiU, ERC and CUP) in all regional elections from 2006 to 2017.

The pattern is clear: national identity becomes an increasingly important determinant of the vote. The big change can be observed in the 2012 elections once the conflict has been activated. While in 2010 the probability to choose one of the same secessionist parties by identifying as ‘only Catalan’ is 65 percent, this probability increases to over 90 percent in the

2012 elections. This pattern then holds for the following elections (2015 and 2017), suggesting that the ethnic polarisation is not limited to a single election, but sustained over time. The same finding applies to the category ‘more Catalan’, though to a slightly lesser degree. By contrast, the pattern for dual and Spanish identities follow the opposite pattern.

Whereas the probability of voting for secessionist parties in 2010 for those having a dual identity was close to 40 percent before 2012, it is reduced to 15 percent in 2017. The same follows for those in the category of ‘only/more Spanish’, going from 25 percent probability in

2010 to a 2 percent in 2017. If we consider the weight of national identity on voting behaviour by party-elections, the results are consistent (see figure A.3 in the appendix).

Notably, CiU would have suffered significant losses amongst those in the categories of

‘Only/More Spanish’ and ‘Dual’ in the 2012 elections. At the same time, Ciudadanos would have started gaining the vote for those categories, steadily increasing in 2015 and 2017.

These results largely confirm hypothesis 3 by showing that there has been a process of ethnic alignment in Catalonia since 2012, thus also proving the consistency of the findings of

Serrano and Bonillo (2017) at the individual level.

- 157 -

Figure 6.4 Predicted Probabilities of voting for a pro-secessionist party (CiU, ERC and CUP, or coalition JxSi/JxCat) by national identity in different elections.

Notes: Controls include: Left-Right scale, gender, age, education, province and size of municipality. Regression results available upon request. Source: CIS post-electoral studies (2660; 2857; 2970; 3113; 3202).

Stepwise Mobilisation

The fourth and final hypothesis (H4) raised the question of political activation. The fact that electoral turnout increased in each election, especially from 2010 (58.67%) to 2012

(67.76%), raises the question “as to what extent Catalan nationalism would have appealed to the population born in the rest of Spain, emphasizing welfare and civic aspects” (Serrano and

Bonillo 2017: 375). In order to test H4, I check to what extent being in favour of “a State that recognises the possibility of autonomous communities to become an independent state” predicts the probability of having abstained in each of the elections9. Figure 6.5 illustrates the

9 Only those who could vote but did not do so are included as in the sample as ‘abstainers’.

- 158 -

predicted probabilities of abstaining in the election as a function of support for independence.

Figure 6.5 Predicted Probabilities of abstaining by preferences over territorial organisation in different elections.

Notes: Controls include: Left-Right scale, gender, age, education, province and size of municipality. Regression results available upon request. Source: CIS post-electoral studies (2660; 2857; 2970; 3113; 3202).

Overall, there is, as expected given the higher turnout, a significant decline of the probability of abstaining for both groups. However, we observe that in the 2012 elections the decline is significantly more pronounced for the pro-independence camp. In the 2012 elections, the first in which the mainstream right party CiU openly presented itself as pro-secessionist, we observe that the supporters of independence are less likely to abstain, although there is also a significant decline in the probability of abstaining for the non-independence camp as well.

This pattern is then corrected in the 2015 and even more so in the 2017 elections, when there is no longer any significant difference between the two.

- 159 -

Further evidence of this two-step process can be observed by looking at a specific group of respondents who declared to have voted in the election, but who claim to rarely or never do so. I shall call this group the politically ‘activated’. Comparing this group to those who declare to always vote and indeed voted in the elections, which I shall call the politically

‘active’, points in the same direction. In the 2015 and 2017 elections, the probability of being in the political ‘activated’ group significantly grows amongst those who did not vote for a pro-secessionist party. In other words, the politically ‘activated’, are, on average, less likely to have voted for a pro-independence party in the 2015 and 2017 elections (see figure A5 in the appendix). Not surprisingly, the same applies when instead of analysing the vote for secessionist parties we look at national identity (see figure A6 in the appendix). In the latter two elections (2015 and 2017), a higher proportion of respondents who claim to have voted but rarely or never did in previous elections fall within the category of dual and Spanish identities. This effect is not observed in the 2012 elections.

The results of H4 point to a two-step process as a consequence of ethnic polarisation: it is those who favoured secessionism who got initially activated the most, but in the elections that followed those opposing it caught up. The fact that the latter group took longer to active as a response to ethnic polarisation can be explained by the fact that those supporting secessionism led the initiative and the political opposition took longer to consolidate (i.e. the consolidation of Ciudadanos, see Teruel and Barrio (2016)), but also by the fact that national identities, and thus support for secessionism is correlated with education (see Figure A7 in the appendix). As we know from previous literature, disadvantaged individuals are more difficult to mobilize (Persson 2015). As it turns out, they were more numerous in the anti- secessionist camp.

- 160 -

Conclusion

The activation of the secessionist conflict in Catalonia in the aftermath of the Great

Recession raises important questions about representation, conflict and party competition dynamics to which this article has aimed to contribute. Theoretically, this chapter has advanced the understanding on how mainstream parties can engage in strategies of conflict displacement when faced with adverse political circumstances. Empirically, it has shown how such strategies can shape the public debate and ultimately transform political conflict as a result.

The new political scenario in Catalonia after the procés towards independence is one of high polarisation over national identity in which the fundamental idea of the state is at stake. This process has resulted in an extraordinary degree of formal and informal political participation, as seen by increasing turnout from both camps but also in the mobilisations in the protest arena. While normatively speaking higher participation might be conceived as desirable, it has also been accompanied by extreme polarisation, which has increased as the conflict developed.

While the secessionist conflict that activated in Catalonia is closely related to identitarian divides —and a much deeper historical cleavage over nation state-building—, it does not mean that it is detached from the Great Recession. This is because, as it also seems to be the case in several other European countries, economic grievances are translated into political grievances, often in the form of conflicts over identity. In the Spanish case, the economic crisis exposed the incapacity of mainstream parties to offer an adequate programmatic response to the management of the crisis. This produced a profound crisis a representation and a potential for alternative narratives to fill the new political void.

Podemos, as advanced in the previous chapter, managed to combine opposition to austerity with a demand for new politics. In Catalonia, an alternative narrative proposed the creation of

- 161 -

a new state as a form of democratic renewal and reversal of austerity. This narrative became extended once the mainstream party CiU saw an opportunity to avoid the risk of political loss by embracing secessionism.

One important implication stems from this analysis: more attention ought to be paid to strategies of actors in shaping the public debate for understanding the structures of conflict.

Political conflict does not transform merely due to the aggregation of preferences from the demand-side of politics. It is political actors that shape the agenda and may activate or de- activate conflict depending on their interests. After all, the power to shape the political agenda is ultimately in the hands of the political elites, and these have at their disposal a range of possibilities to influence the public debate.

- 162 -

Chapter 7. Conclusion: Towards a New Structure of Conflict in Southern Europe?

The aim of this dissertation was to better comprehend the political transformations and political restructuring that the Great Recession triggered in southern Europe in the shadow of the Great Recession (2008—). For that purpose, two comparative analyses and three case studies analysed different aspects of the origins and consequences of such transformations.

Before I lay out some of the conclusions that this study has to offer, it is worth highlighting two important limitations. The first of them is the magnitude of the phenomenon studied. The

“political consequences” of something as colossal as an economic shock that affected several continents, especially given the many dimensions of politics and their interplay, is too much for any dissertation to tackle. Going into the specifics of the events, actors, strategies and their motivations as the crisis unfolded in comparative perspective for the four countries under study —i.e. Greece, Italy, Portugal and Spain— would have been preferable, but also unrealistic. While trying to provide a bird-eye’s view comparative framework, I decided to focus on Spain as a paradigmatic case where the transformations of the party system were significant and not limited to the emergence of a populist left party as the theory would predict. This approach, in my view, allows to expose how external economic shocks can interact with domestic conflicts, providing some insight into the dynamics of politics in a real-world setting and allowing for a certain degree of extrapolation to the experiences of other countries facing similar contexts.

The second limitation is that the subject of study deals with very recent phenomena, to the extent that some of the chapters were written as the events were unfolding. Even if some of the chapters have been somewhat updated a posteriori, it is impossible to disentangle where the political consequences of the Great Recession end exactly. To the misfortune of

- 163 -

analytical clarity, this is a path-dependence political world in constant flux. The cut-off point in the analysis of this dissertation has thus depended more on time limitations than on any solid theoretical reason. It could be argued that we are only now starting to see the real political consequences of the Great Recession, or at least an important part of them. Future research should pay close attention to the emergence and success of the radical right in these countries and the possible consolidation of a new cultural cleavage as established in north- western Eruope.

Despite these limitations, each of the chapters has some offerings, which taken cumulatively, give us some ideas on the effect of the economic crisis in southern European politics. Chapter 2 compared the economic and political origins and consequences of the crisis in Greece and Spain and argued that the ill-conceived institutional structure of the EU, together with prior political domestic developments, created a socio-economic environment in which the economic crisis could only have the disruptive political consequences that we witnessed. It was also shown that, while austerity packages were much harsher in Greece than in Spain, both countries suffered dramatic economic and social consequences, which triggered a severe distrust towards political institutions and bred the ground for the restructuring of their party systems.

Chapter 3 took a bird-eye perspective of the macro-political transformations in the four southern European countries —Greece, Italy, Portugal, and Spain— and showed that political conflict is structured around economic and political issues at the European and domestic levels. These transformations were driven by new parties that combined an anti- austerity platform with promises of democratic regeneration. Moreover, this structure was found to differ from the structure found to shape the Northwest of Europe, where the driving forces are the populist radical right and a new cultural cleavage is structuring the political space.

- 164 -

Chapter 4 focused on Spain to trace the evolution of domestic political conflict and analysed the specific types of issues that structured political competition before and after the crisis period in more detail. In the years previous to the crisis, cultural issues and security concerns dominated the public debate. Once the crisis struck, the issues in the political debate changed dramatically. Whereas economic issues dominated during the first stages of the crisis, political issues concerning democratic renewal and the nation-building dimension of

Spanish politics rapidly became more prominent in later stages. An important takeaway from this chapter is that economic grievances can be rapidly channelled through other types of political conflict that do not necessarily involve economic issues.

Chapter 5 focused on the emergence of the two new parties in the Spanish case —

Podemos and Ciudadanos— and tested several hypotheses about the underlying factors that led to their emergence by analysing the characteristics of their electorate. It was argued that a crisis of representation (i.e. political grievances) is a much more important determinant of the vote for these new parties than traditional issues. Moreover, it was shown that despite the emergence of the new parties, the structure of conflict remains unidimensional in the demand side. Finally, it was illustrated how a generational divide that cuts across the Spanish electorate also interacted with the dissatisfaction that drove the vote for both Podemos and

Ciudadanos: voters of these parties represent a younger and dissatisfied share of the electorate on opposite sides of the political spectrum.

Finally, Chapter 6 focused on the case of Catalonia to explore one of the main issues that dominated the Spanish political agenda in the later stages of the crisis, namely the nation-building conflict. This last chapter illustrates that the crisis was not limited to economic consequences but that the activation of conflict was strongly mediated by the interests of political actors. More specifically, the chapter discusses displacement strategies that parties can follow when they face the risk of losing power, as well as the consequences

- 165 -

that the activation had in the Catalan experience. It is shown that the electorate underwent a process of ethnic polarisation and a step-wise increase in participation by both camps. This chapter explores in more detail the idea raised in chapter 4 on how political grievances resulting from an economic shock can be politicised in different ways depending on the historical and political legacies of the territory.

Parenthesis towards convergence or exceptionalism?

Analysing ongoing political phenomena is a risky endeavour. By the time of writing, a new radical right party (VOX) has emerged in the regional Andalusian elections in Spain with the prospects of gaining representation at the national parliament in the next general elections.

Speculating on the future of political conflict in Spain, one could argue that the emergence of a such a party in Spain is part of a ‘catching-up process’ of south European politics to the

North-Western European structure of conflict. Following this perspective, the political transformations during the Great Recession would have merely constituted a parenthesis towards a more stable cultural divide represented by two poles: the populist radical right and the green or alternative parties. However, an argument could also be made for the exceptionality of the southern region for several reasons. First, the radical right parties in these four countries such as Golden Dawn represent more traditional far right parties rather than the ‘new right’ that is gaining strength in NWE. It is too early to tell for the Spanish case, though the leaders of the new party are radical factions of the mainstream right and there are no explicit rejections of its fascist past, quite the contrary. While Italy seems the

‘best-fitting case’ with the Lega’s transformation from a regionalist party to PRR, Portugal has no populist radical right in sight.

On the green pole, the equivalence is not so straightforward either. While Podemos,

Syriza, and the Bloco de Esquerda in Portugal might eventually become the functional

- 166 -

equivalent of the north-European green parties, they are for now different animals, mixing elements of the new left with the traditional radical left. The story is even less clear for the

M5S in Italy, which has evolved into a centre populist party, progressively abandoning any leftist tendencies it might have had when it first formed in 2013 and even adopting anti- immigration stances when necessary. For now, it appears as though the opportunity structures that opened for these parties by combining anti-austerity claims with a new way of doing politics have allowed them to consolidate a new left identity beyond green parties, which in some cases such as Podemos (Equo) are embodied within these coalitions.

Has the Great Recession constituted a parenthesis in the convergence of southern

European politics towards the North, or is southern Europe exceptional and following its own path? To this question we can only speculate. It is reasonable to think that, once the crisis of representation has passed and new parties consolidate, these parties will be forced to better define themselves ideologically in more rooted structures of conflict. In other words, it is likely that the exceptionality of the crisis allowed new left parties such as Podemos and

Syriza to capitalise on the discontent and feeling of detachment from politics, but once the storm is over, these parties will need to find ways to reposition themselves in a new structure of conflict.

- 167 -

- 168 -

Bibliography Abedi, Arim (2002). ‘Challenges to established parties: The effects of party system features on the electoral fortunes of anti-political-establishment parties’, European Journal of Political Research, 41:4, 551–583. Afonso, Alexandre, Sotirios Zartaloudis, and Yannis Papadopoulos (2014). ‘How party linkages shape austerity politics: clientelism and fiscal adjustment in Greece and Portugal during the eurozone crisis’, Journal of European Public Policy, 22:3, 315– 334. Akkerman, Tjitske (2015). ‘Immigration policy and electoral competition in Western Europe: A fine-grained analysis of party positions over the past two decades’, Party Politics, 21:1, 54–67. Alesina, Alberto, and Silvia Ardagna (2010). ‘Large Changes in Fiscal Policy: Taxes versus Spending’, Tax Policy and the Economy, 24:1, 35–68. Alesina, Alberto, and Roberto Perotti (1997). ‘Fiscal Adjustments in OECD Countries: Composition and Macroeconomic Effects’, Staff Papers (International Monetary Fund), 44:2, 210–248. Alonso, Sonia, and Cristóbal Rovira Kaltwasser (2015). ‘Spain: No Country for the Populist Radical Right?’, South European Society and Politics, 20:1, 21–45. Altiparmakis, Argyrios (2018). ‘Greece – Punctuated equilibrium: Restructuring Greek party politics’, in Swen Hutter and Hanspeter Kriesi (eds.), Restructuring European Party Politics in Times of Crisis, vol. Cambridg: Cambridge University Press. Altiparmakis, Argyrios, and Jasmine Lorenzini (2018). ‘Disclaiming national representatives: Protest waves in Southern Europe during the crisis’, Party Politics, 24:1, 78–89. Andersen, Jørgen Juel, Jon H. Fiva, and Gisle James Natvik (2014). ‘Voting when the stakes are high’, Journal of Public Economics, 110, 157–166. Anderson, Christopher J. (2000). ‘Economic voting and political context: a comparative perspective’, Electoral Studies, 19:2, 151–170. Anduiza, Eva, Camilo Cristancho, and José M. Sabucedo (2014). ‘Mobilization through online social networks: the political protest of the indignados in Spain’, Information, Communication & Society, 17:6, 750–764. Anon ‘The sick man of the euro’. The Economist, June 3 http://www.economist.com/node/209559 (Accessed October 26, 2015). Ares, Macarena, and Enrique Hernández (2017). ‘The corrosive effect of corruption on trust in politicians: Evidence from a natural experiment’, Research & Politics, 4:2, 1–8. Astudillo, J. (2009). ‘Le sconfitte di Rajoy: la destra dopo Aznar’, Bosco and Sánchez- Cuenca, eds., La Spagna di Zapatero,, 67–86. Balaguer, Jaime, and Alberto Sanz (2010). ‘La segunda ola de la crispación: competición y polarización en la VIII legislatura’, in Jose Ramon Montero and Ignacio Lago (eds.), Elecciones Generales 2008, vol., Elecciones. Madrid: Centro de Investigaciones Sociologicas, 61–92. Balcells, Laia, José Fernández-Albertos, and Alexander Kuo (2015). ‘Preferences for Inter- Regional Redistribution’, Comparative Political Studies,, 0010414015581681. Balfour, Sebastian (2005). ‘The reinvention of Spanish conservatism: the Popular Party since 1989’, in Sebastian Balfour (ed.), The politics of contemporary Spain., vol. London: Routledge. Baras, Montserrat, Oscar Barberà, Astrid Barrio, and Juan Rodríguez-Teruel (2015). ‘How Does May’s Law Work in Multi-dimensional Competition? Intra-party Opinion Structure in Catalonia’, World Political Science, 11:2

- 169 -

https://www.degruyter.com/view/j/wps.2015.11.issue-2/wps-2015-1005/wps-2015- 1005.xml (Accessed October 2, 2018). Barrio, Astrid, Oscar Barberà, and Juan Rodríguez-Teruel (2018). ‘“Spain steals from us!” The “populist drift” of Catalan regionalism’, Comparative European Politics, http://link.springer.com/10.1057/s41295-018-0140-3 (Accessed October 17, 2018). Barrio, Astrid, and Juan Rodríguez-Teruel (2016). ‘Reducing the gap between leaders and voters? Elite polarization, outbidding competition, and the rise of secessionism in Catalonia’, Ethnic and Racial Studies,, 1–19. Barrio, Astrid, and Juan Rodríguez-Teruel (2017). ‘Reducing the gap between leaders and voters? Elite polarization, outbidding competition, and the rise of secessionism in Catalonia’, Ethnic and Racial Studies, 40:10, 1776–1794. Bartolini, Stefano, and P. Mair (1990). Identity, Competition, and Electoral Availability: The Stabilisation of European Electorates 1885-1985. Cambridge University Press. Basta, Karlo (2017). ‘The Social Construction of Transformative Political Events’, Comparative Political Studies, http://journals.sagepub.com/doi/abs/10.1177/0010414017740601?journalCode=cpsa# articleCitationDownloadContainer (Accessed November 16, 2017). Bayoumi, Tamim, and Barry Eichengreen (1997). ‘Ever closer to heaven? An optimum- currency-area index for European countries’, European Economic Review, 41:3–5, 761–770. Beramendi, Pablo, Silja Häusermann, Herbert Kitschelt, and Hanspeter Kriesi, eds. (2015a). The Politics of Advanced Capitalism. New York: Cambridge University Press. Beramendi, Pablo, Silja Häusermann, Herbert Kitschelt, and Hanspeter Kriesi (2015b). ‘Introduction: The Politics of Advanced Capitalism’, in Pablo Beramendi,Silja Häusermann,Herbert Kitschelt,and Hanspeter Kriesi (eds.), The Politics of Advanced Capitalism, vol. New York: Cambridge University Press, 1–64. Bermeo, Nancy (1997). ‘Myths of Moderation: Confrontation and Conflict during Democratic Transitions’, Comparative Politics, 29:3, 305. Biorcio, Roberto (2014). ‘The reasons for the success and transformations of the 5 Star Movement’, Contemporary Italian Politics, 6:1, 37–53. Biorcio, Roberto (2015). Il populismo nella politica italiana : da Bossi a Berlusconi, da Grillo a Renzi. Milano: Mimesis. Bisgaard, Martin, and Rune Slothuus (2018). ‘Partisan Elite Influence in a Real-World Setting’, European University Institute, Florence. Blyth, Mark (2013). Austerity: The History of a Dangerous Idea. Oxford ; New York: Oxford University Press. Bobba, Giuliano, and Duncan McDonnell (2015). ‘Italy – a strong and enduring market for populism’, in Hanspeter Kriesi and Takis Pappas (eds.), European Populism in the Shadow of the Great Recession, vol. Colchester, U.K.: ECPR Press, 159–174. Bofinger, Peter, Lars Feld, Christoph Schmidt, Isabel Schnabel, and Volker Wieland (2015). ‘Consequences of the Greek Crisis for a More Stable Euro Area’, , Wiesbaden: German Council of Economic Experts. Bonet, Eduard, Santiago Pérez-Nievas, and María José Hierro (2010). ‘España en las urnas: Territorialization del voto e identidad nacional en las elecciones de 2008’, in Jose Ramon Montero and Ignacio Lago (eds.), Elecciones Generales 2008, vol., Elecciones. Madrid: Centro de Investigaciones Sociologicas. Borghetto, Enrico, and Federico Russo (2018). ‘From agenda setters to agenda takers? The determinants of party issue attention in times of crisis’, Party Politics, 24:1, 65–77. Bornschier, Simon (2010). Cleavage politics and the populist right: the new cultural conflict in Western Europe. Philadelphia: Temple University Press.

- 170 -

Bosch, Agusti, and Iván M Durán (2017). ‘How does economic crisis impel emerging parties on the road to elections? The case of the Spanish Podemos and Ciudadanos’, Party Politics,, 135406881771022. Bosco, Anna, and Susannah Verney (2012). ‘Electoral Epidemic: The Political Cost of Economic Crisis in Southern Europe, 2010–11’, South European Society and Politics, 17:2, 129–154. Botti, Alfonso (2013). ‘From Opposition to Government: The Popular Party of Mariano Rajoy’, in Bonnie N. Field in political science and Alfonso Botti (eds.), Politics and Society in Contemporary Spain, vol., Europe in Transition: The NYU European Studies Series. Palgrave Macmillan US, 41–59 http://link.springer.com/chapter/10.1057/9781137306623_3 (Accessed December 5, 2016). Bremer, Björn (2018). ‘The missing left? Economic crisis and the programmatic response of social democratic parties in Europe’, Party Politics, 24:1, 23–38. Buck, Tobias ‘Rajoy promises jobs and growth as fringe parties gain ground’. Financial Times, February 24 http://www.ft.com/intl/cms/s/0/5876fd4c-bc30-11e4-b6ec- 00144feab7de.html#axzz3pHeBhojo (Accessed October 23, 2015). Cainzos, Miguel, and Carmen Voces (2015). ‘Paro y voto: ¿afecta al voto la experiencia de desempleo?’, Revista de Estudios Políticos,168, 115–150. Calvo, Kerman (2013). ‘Fighting for a voice: The Spanish 15-M/Indignados movement’, in Cristina Flesher Fominaya and Laurence Cox (eds.), Understanding European Movements: New Social Movements, Global Justice Struggles, Anti-Austerity Protest, vol. Routledge. Calvo, Kerman, Teresa Gómez-Pastrana, and Luis Mena (2011). ‘Movimiento 15M: ¿quiénes son y qué reivindican?’, Zoom Político, Laboratorio de Alternativas. Calvo, Kerman, Alvaro Martinez, and Jose Ramon Montero (2010). ‘Devotos y votantes: El peso del factor religioso en las elecciones generales’, in José Ramón Montero and Ignacio Lago (eds.), Elecciones Generales 2008, vol. Madrid: Centro de Investigaciones Sociológicas, 235–268. Castañeda, Ernesto (2012). ‘The Indignados of Spain: A Precedent to Occupy Wall Street’, Social Movement Studies, 11:3–4, 309–319. Castro, Carles (2008). Relato electoral de España (1977-2077). Barcelona: Institut de Ciencies Politiques i Socials. Ceccarini, Luigi, and Fabio Bordignon (2016). ‘The five stars continue to shine: the consolidation of Grillo’s “movement party” in Italy’, Contemporary Italian Politics, 8:2, 131–159. Colomer, Josep M. (1998). ‘The Spanish “state of autonomies”: Non‐institutional federalism’, West European Politics, 21:4, 40–52. Connelly, Roxanne, Vernon Gayle, and Paul S. Lambert (2016). ‘A Review of occupation- based social classifications for social survey research’, Methodological Innovations, 9, 2059799116638003. Conti, Nicolò, Swen Hutter, and Kyriaki Nanou (2018). ‘Party competition and political representation in crisis: An introductory note’, Party Politics, 24:1, 3–9. Conti, Nicolò, and Vincenzo Memoli (2015). ‘The Emergence of a New Party in the Italian Party System: Rise and Fortunes of the Five Star Movement’, West European Politics, 38:3, 516–534. Copelovitch, Mark, Jeffry Frieden, and Stefanie Walter (2016). ‘The Political Economy of the Euro Crisis’, Comparative Political Studies, 49:7, 811–840. Dahl, Robert A (1956). A preface to democratic theory. University of Chicago Press. Dahl, Robert A. (2000). ‘A Democratic Paradox?’, Political Science Quarterly, 115:1, 35–40.

- 171 -

Dalton, Russell J., Scott C. Flanagan, and Paul Allen Beck, eds. (1984). Electoral Change in Advanced Industrial Democracies: Realignment or Dealignment? Princeton, N.J: Princeton University Press. De Grauwe, Paul (2012). Economics of monetary union. Ninth edition. Oxford: Oxford University Press. De Grauwe, Paul (2013). ‘The Political Economy of the Euro’, Annual Review of Political Science, 16:1, 153–170. De Grauwe, Paul, and Yuemei Ji (2014). ‘The Future of the Eurozone’, The Manchester School, 82, 15–34. De Vries, Catherine E., and Gary Marks (2012). ‘The struggle over dimensionality: A note on theory and empirics’, European Union Politics, 13:2, 185–193. Della Porta, Donatella, Francis O’Connor, Martín Portos, and Anna Subirats Ribas (2017). Social movements and referendums from below: direct democracy in the neoliberal crisis. Diez Medrano, Juan (2003). Framing Europe. Attitudes to European integration in Germany, Spain, and the United Kingdom. Princeton: Princeton University Press. Díez-Medrano, Juan (2003). Framing Europe: attitudes to European integration in Germany, Spain, and the United Kingdom. Princeton, N.J: Princeton University Press. Dinas, Elias (2012). ‘Left and Right in the Basque Country and Catalonia: The Meaning of Ideology in a Nationalist Context’, South European Society and Politics, 17:3, 467– 485. Eichengreen, Barry (2012). ‘European monetary integration with benefit of hindsight’, JCMS: Journal of Common Market Studies, 50:s1, 123–136. Encarnación, Omar Guillermo (2004). ‘The politics of immigration: Why Spain is different’, Mediterranean Quarterly, 15:4, 167–185. Evans, Geoffrey, and Nan Dirk de Graaf (2013). Political Choice Matters: Explaining the Strength of Class and Religious Cleavages in Cross-National Perspective. OUP Oxford. Evans, Geoffrey, and Mark Pickup (2010). ‘Reversing the Causal Arrow: The Political Conditioning of Economic Perceptions in the 2000–2004 U.S. Presidential Election Cycle’, The Journal of Politics, 72:4, 1236–1251. Evans, Geoffrey, and James Tilley (2017). The New Politics of Class: The Political Exclusion of the British Working Class. Oxford, New York: Oxford University Press. Fabre, E. (2010). ‘Measuring party organization: The vertical dimension of the multi-level organization of state-wide parties in Spain and the UK’, Party Politics, 17:3, 343– 363. Featherstone, Kevin (2011). ‘The JCMS Annual Lecture: The Greek Sovereign Debt Crisis and EMU: A Failing State in a Skewed Regime’, JCMS: Journal of Common Market Studies, 49:2, 193–217. Fernandes, Tiago (2015). ‘Rethinking pathways to democracy: civil society in Portugal and Spain, 1960s–2000s’, Democratization, 22:6, 1074–1104. Fernández-Albertos, José (2015a). Los votantes de Podemos : del partido de los indignados al partido de los excluidos. Madrid: Catarata. Fernández-Albertos, José (2015b). Los votantes de Podemos: del partido de los indignados al partido de los excluidos. Los Libros de la Catarata. Fernández-Albertos, José (2017). ‘Nuevos partidos, ¿nuevos ejes de competición política?’, Piedras de Papel, http://www.eldiario.es/piedrasdepapel/Nuevos-partidos-nuevos- competicion-politica_6_629997015.html (Accessed April 17, 2017). Fernández-Albertos, José, and Dulce Manzano (2012a). ¿Quién apoya el Estado del Bienestar? Madrid: More izquierdas - Fundación Alternativas

- 172 -

http://www.fundacionalternativas.org/las-publicaciones/libros/quien-apoya-el-estado- del-bienestar (Accessed November 18, 2016). Fernández-Albertos, José, and Dulce Manzano (2012b). ‘The Lack of Partisan Conflict over the Welfare State in Spain’, South European Society and Politics, 17:3, 427–447. Fernández-Albertos, José, and Dulce Manzano (2016). ‘Dualism and support for the welfare state’, Comparative European Politics, 14:3, 349–375. Fernández-Vázquez, Pablo, Pablo Barber?, and Gonzalo Rivero (2016). ‘Rooting Out Corruption or Rooting for Corruption? The Heterogeneous Electoral Consequences of Scandals’, Political Science Research and Methods, 4:02, 379–397. Fernández-Villaverde, Jesús, Luis Garicano, and Tano Santos (2013). ‘Political Credit Cycles: The Case of the Eurozone’, Journal of Economic Perspectives, 27:3, 145– 166. Ferreira, Frederico, and Mariana S. Mendes (2017). ‘Portugal: A stability tale or a party- induced transformation?’, Presented at the workshop ‘Southern Europe in Crisis: A Comparative Perspective’, Florence, Italy: European University Institute. Ferrera, Maurizio (1996). ‘The’Southern model’of welfare in social Europe’, Journal of European social policy,6, 17–37. Ferrera, Maurizio, ed. (2005). Welfare state reform in Southern Europe: fighting poverty and social exclusion in Italy, Spain, Portugal andGreece. London ; New York: Routledge. Field, Bonnie N. (2009). ‘A “Second Transition” in Spain? Policy, Institutions and Interparty Politics under Zapatero (2004–8)’, South European Society and Politics, 14:4, 379– 397. Field, Bonnie N., and Alfonso Botti, eds. (2013). ‘Economic Reforms and the Labor Market: Zapatero’s Endless Period in the Wilderness’, in Politics and Society in Contemporary Spain, vol. Palgrave Macmillan, 101–121 http://www.palgraveconnect.com/doifinder/10.1057/9781137306623 (Accessed February 25, 2016). Franklin, Mark N, Thomas T Mackie, and Henry Valen (1992). Electoral change: Responses to evolving social and attitudinal structures in Western countries. Cambridge University Press. Galais, Carol (2014). ‘Don’t Vote for Them: The Effects of the Spanish Indignant Movement on Attitudes about Voting’, Journal of Elections, Public Opinion and Parties, 24:3, 334–350. Galindo, Jorge et al. (2015). Podemos: la cuadratura del círculo. Madrid: ENDEBATE. García-Albacete, Gema, Javier Lorente, and Irene Martín (2016a). ‘How does the Spanish crisis generation relates to politics?’, Thijssen, Peter;Siongers, Jessy;Van Laer, Jeroen;Haers, Jacques; Mels, Sara, 50–72 http://books.google.com/books?hl=en&lr=&id=w6s0CwAAQBAJ&oi=fnd&pg=PA5 0&dq=info:H8HMzFA3Y5AJ:scholar.google.com&ots=jUzHQbfWV8&sig=Yz4DY C-igF2BsFvgTA8qYv84aO0 (Accessed June 20, 2017). García-Albacete, Gema, Javier Lorente, and Irene Martín (2016b). ‘How does the “Spanish crisis generation” relate to politics?’, in Peter Thijssen,Jessy Siongers,Jeroen Van Laer,Jacques Haers,and Sara Mels (eds.), Political Engagement of the Young in Europe: Youth in the Crucible, vol. Routledge. Gattinara, Pietro Castelli (2016). The Politics of Migration in Italy: Perspectives on local debates and party competition. 1 edizione. London ; New York, NY: Routledge. Gemenis, Kostas, and Roula Nezi (2015). ‘Government–Opposition Dynamics during the Economic Crisis in Greece’, The Journal of Legislative Studies, 21:1, 14–34. Giavazzi, Francesco, and Luigi Spaventa (2011). ‘Why the Current Account Matters in a Monetary Union’, in Miroslav Beblavy,David Cobham,and L’udovit Odor (eds.), The

- 173 -

Euro Area and The Financial Crisis, vol. Cambridge: Cambridge University Press, 59–80. Golder, Matt (2003). ‘Explaining variation in the success of extreme right parties in Western Europe’, Comparative Political Studies, 36:4, 432–466. Guibernau, Montserrat (2014). ‘Prospects for an Independent Catalonia’, International Journal of Politics, Culture, and Society, 27:1, 5–23. Guinjoan, Marc, and Toni Rodon (2016). ‘A Scrutiny of the Linz-Moreno Question’, Publius: The Journal of Federalism, 46:1, 128–142. Gunther, Richard (2005). ‘Parties and Electoral Behavior in Southern Europe’, Comparative Politics, 37:3, 253–275. Gunther, Richard, and Jose Ramon Montero (1994). ‘Los anclajes del partidismo: un análisis comparado del comportamiento electoral en cuatro democracias del sur de Europa’, in Pilar del Castillo (ed.), Comportamiento político y electoral, vol. Madrid: Centro de Investigaciones Sociologicas. Gunther, Richard, Giacomo Sani, and Goldie Shabad (1986). ‘El pasado como prólogo’, in El sistema de partidos en España. Génesis y evolución, vol. Madrid: Centro de Investigaciones Sociologicas, 13–43. Hall, Peter A. (2012). ‘The Economics and Politics of the Euro Crisis’, German Politics, 21:4, 355–371. Hansen, Randall, and Joshua C. Gordon (2014). ‘Deficits, Democracy, and Demographics: Europe’s Three Crises’, West European Politics, 37:6, 1199–1222. Harmel, Robert, and John D. Robertson (1985). ‘Formation and Success of New Parties A Cross-National Analysis’, International Political Science Review, 6:4, 501–523. Harmel, Robert, and Lars Svasand (1997). ‘The Influence of New Parties on Old Parties’ Platforms The Cases of the Progress Parties and Conservative Parties of Denmark and Norway’, Party Politics, 3:3, 315–340. Hellwig, T., and D. Samuels (2007). ‘Voting in Open Economies: The Electoral Consequences of Globalization’, Comparative Political Studies, 40:3, 283–306. Hernández, Enrique, and Hanspeter Kriesi (2015). ‘The electoral consequences of the financial and economic crisis in Europe’, European Journal of Political Research,, Advanced online publication. Hernández, Enrique, and Hanspeter Kriesi (2016). ‘The electoral consequences of the financial and economic crisis in Europe’, European Journal of Political Research, 55:2, 203–224. Hino, Airo (2012). New Challenger Parties in Western Europe: A Comparative Analysis. Routledge. Hobolt, Sara B., and James Tilley (2014). Blaming Europe?: Responsibility Without Accountability in the European Union. New York, NY: OUP Oxford. Hobolt, Sara B., and James Tilley (2015). ‘Fleeing the Centre: The Rise of Challenger Parties in the Aftermath of the Euro Crisis’, http://eeshomepage.net/wp- content/uploads/2015/11/HoboltTilley_FleeingTheCentre.pdf (Accessed September 23, 2016). Hobolt, Sara B., and James Tilley (2016). ‘Fleeing the centre: the rise of challenger parties in the aftermath of the euro crisis’, West European Politics, 39:5, 971–991. Hobolt, Sara B., and Catherine de Vries (2016). ‘Turning against the Union? The impact of the crisis on the Eurosceptic vote in the 2014 European Parliament elections’, Electoral Studies, 44, 504–514. Hobolt, Sara B., and Catherine E. de Vries (2015). ‘Issue Entrepreneurship and Multiparty Competition’, Comparative Political Studies, 48:9, 1159–1185.

- 174 -

Hooghe, L., G. Marks, and C. J. Wilson (2002). ‘Does Left/Right Structure Party Positions on European Integration?’, Comparative Political Studies, 35:8, 965–989. Hooghe, Liesbet, and Gary Marks (2017). ‘Cleavage theory meets Europe’s crises: Lipset, Rokkan, and the transnational cleavage’, Journal of European Public Policy,, 1–27. Hooghe, Marc, and Ruth Dassonneville (2016). ‘A Spiral of Distrust: A Panel Study on the Relation between Political Distrust and Protest Voting in Belgium’, Government and Opposition, AOP, 1–27. Hopkin, Jonathan (2001a). ‘A “southern model” of electoral mobilisation? Clientelism and electoral politics in Spain’, West European Politics, 24:1, 115–136. Hopkin, Jonathan (2001b). ‘A “southern model” of electoral mobilisation? Clientelism and electoral politics in Spain’, West European Politics, 24:1, 115–136. Hopkin, Jonathan (2002). ‘The emergence and convergence of the cartel party: parties, State and economy in Southern Europe’,. Hopkin, Jonathan (2005). ‘Spain: Proportional Representation with Majoritarian Outcomes’, in Michael Gallagher and Paul Mitchell (eds.), The politics of electoral systems, vol. Oxford: Oxford University Press http://www.oxfordscholarship.com.ezproxy.eui.eu/view/10.1093/0199257566.001.00 01/acprof-9780199257560-chapter-18 (Accessed November 22, 2016). Hopkin, Jonathan (2009). ‘Party Matters. Devolution and Party Politics in Britain and Spain’, Party Politics, 15:2, 179–198. Hopkin, Jonathan (2012). ‘Clientelism, corruption and political cartels: Southen Europe’, in Thomas Christiansen and Christine Neuhold (eds.), International Handbook on Informal Governance, vol. Cheltenham: Edward Elgar, 198–215. Hughes, Neil (2011). ‘“Young People Took to the Streets and all of a Sudden all of the Political Parties Got Old”: The 15M Movement in Spain’, Social Movement Studies, 10:4, 407–413. Hutter, Swen (2014). Protesting Culture and Economics in Western Europe: New Cleavages in Left and Right Politics. University of Minnesota Press. Hutter, Swen, Edgar Grande, and Hanspeter Kriesi (2016). Politicising Europe: Integration and Mass Politics. Cambridge: Cambridge University Press. Hutter, Swen, Hanspeter Kriesi, and Guillem Vidal (2015). ‘Restructuring the Southern European party systems in the Great Recession’, European University Institute, Florence. Hutter, Swen, Hanspeter Kriesi, and Guillem Vidal (2018). ‘Old versus new politics: The political spaces in Southern Europe in times of crises’, Party Politics, 24:1, 10–22. Inglehart, Ronald (1990). Culture shift in advanced industrial society. Princeton, New Jersey: Princeton University Press. Inglehart, Ronald (1997). Modernization and postmodernization: Cultural, economic, and political change in 43 societies. Princeton University Press. Inglehart, Ronald (2015). The silent revolution: Changing values and political styles among Western publics. Princeton University Press. Inglehart, Ronald, and Scott C. Flanagan (1987). ‘Value Change in Industrial Societies’, The American Political Science Review, 81:4, 1289. Jackman, Robert W, and Karin Volpert (1996). ‘Conditions favouring parties of the extreme right in Western Europe’, British Journal of Political Science, 26:04, 501–521. Jonung, Lars, Eoin Drea, European Commission, and Directorate-General for Economic and Financial Affairs (2009). The euro--it can’t happen, it’s a bad idea, it won’t last US economists on the EMU, 1989-2002. Brussels: European Commission, Directorate- General for Economic and Financial Affairs

- 175 -

http://ec.europa.eu/economy_finance/publications/publication16345_en.pdf (Accessed October 22, 2015). Kaldor, Mary, and Sabine Selchow (2015). Subterranean Politics in Europe. Palgrave Macmillan. Kanellopoulos, Kostas, and Konstantinos Kostopoulos (2014). ‘The major organizations/groups behind the Greek anti-austerity campaign. Repertoires of action and political claims’, http://ecpr.eu/Filestore/PaperProposal/ff897920-1ba0-49c7- 947b-562c8ee29448.pdf (Accessed October 19, 2015). Karyotis, Georgios, and Wolfgang Rüdig (2015). ‘Blame and Punishment? The Electoral Politics of Extreme Austerity in Greece’, Political Studies, 63:1, 2–24. Katsanidou, A., and S. Otjes (2015a). ‘How the European debt crisis reshaped national political space: The case of Greece’, European Union Politics,, Advance online publication. Katsanidou, Alexia, and Simon Otjes (2015b). ‘Mapping the Greek party system after the 2015 elections: how the economy and Europe have merged into a single issue’, EUROPP, http://blogs.lse.ac.uk/europpblog/2015/02/25/mapping-the-greek-party- system-after-the-2015-elections-how-the-economy-and-europe-have-merged-into-a- single-issue/ (Accessed April 4, 2015). Katz, Richard S., and Peter Mair (1995). ‘Changing models of party organization and party democracy the emergence of the cartel party’, Party politics, 1:1, 5–28. Katz, Richard S., and Peter Mair (2009). ‘The Cartel Party Thesis: A Restatement’, Perspectives on Politics, 7:04, 753–766. Key, Valdimer Orlando, and Milton C Cummings (1966a). The responsible electorate. Belknap Press of Harvard University Press. Key, Valdimer Orlando, and Milton C Cummings (1966b). The responsible electorate. Belknap Press of Harvard University Press. Kitschelt, Herbert (1988). ‘Left-Libertarian Parties: Explaining Innovation in Competitive Party Systems’, World Politics, 40:2, 194–234. Kitschelt, Herbert (1994). The transformation of European social democracy. Cambridge New York: Cambridge University Press. Kitschelt, Herbert (1995). ‘Formation of party cleavages in post-communist democracies: Theoretical propositions’, Party politics, 1:4, 447–472. Kitschelt, Herbert (2000). ‘Linkages between Citizens and Politicians in Democratic Polities’, Comparative Political Studies, 33:6–7, 845–879. Kitschelt, Herbert, and Steven I. Wilkinson (2007). ‘Citizen-politician linkages: an introduction’, in Patrons, Clients and Policies, vol. Cambridge University Press, 1– 49. Knigge, Pia (1998). ‘The ecological correlates of right-wing extremism in Western Europe’, European Journal of Political Research, 34:2, 249–279. Krastev, Ivan (2002). ‘The Balkans: Democracy Without Choices’, Journal of Democracy, 13:3, 39–53. Kriesi, Hanspeter (1998). ‘The transformation of cleavage politics The 1997 Stein Rokkan lecture’, European Journal of Political Research, 33:2, 165–185. Kriesi, Hanspeter et al. (2006). ‘Globalization and the transformation of the national political space: Six European countries compared’, European Journal of Political Research, 45:6, 921–956. Kriesi, Hanspeter et al. (2008). West European Politics in the Age of Globalization. Cambridge: Cambridge University Press. Kriesi, Hanspeter (2010). ‘Restructuration of Partisan Politics and the Emergence of a New Cleavage Based on Values’, West European Politics, 33:3, 673–685.

- 176 -

Kriesi, Hanspeter et al. (2012). Political Conflict in Western Europe. Cambridge: Cambridge University Press. Kriesi, Hanspeter (2015). ‘Political Mobilization in Times of Crises: The Relationship between Economic and Political Crises’, in Marco Giugni and Maria Grasso (eds.), Austerity and Protest: Popular Contention in Times of Economic Crisis, vol. Burlington, VT: Ashgate, 19–34. Kriesi, Hanspeter, Laurent Bernhard, and Regula Hänggli (2009). ‘The Politics of Campaigning – Dimensions of Strategic Action’, in Frank Marcinkowski and Barbara Pfetsch (eds.), Politik in der Mediendemokratie, vol. Wiesbaden: VS Verlag für Sozialwissenschaften, 345–365 https://doi.org/10.1007/978-3-531-91728-3_15 (Accessed October 8, 2018). Ladrech, Robert (1989). ‘Social movements and party systems: The French socialist party and new social movements’, West European Politics, 12:3, 262–279. Laffan, Brigid (2016). ‘Europe’s union in crisis: tested and contested’, West European Politics, 39:5, 915–932. Lago, Ignacio, and Jose Ramon Montero (2005). ‘“Todavía no sé quiénes, pero ganaremos”: manipulación política del sistema electoral español’, Zona Abierta, 110/111, 279–348. Lago, Ignacio, José Ramón Montero, and Mariano Torcal (2007). ‘The 2006 Regional Election in Catalonia: Exit, Voice, and Electoral Market Failures’, South European Society and Politics, 12:2, 221–235. Lane, Philip R (2012). ‘The European Sovereign Debt Crisis’, Journal of Economic Perspectives, 26:3, 49–68. Lavezzolo, Sebastián, and Luis Ramiro (2017). ‘Stealth democracy and the support for new and challenger parties’, European Political Science Review, AOP, 1–23. Leeper, Thomas J., and Rune Slothuus (2014). ‘Political Parties, Motivated Reasoning, and Public Opinion Formation: Parties and Motivated Reasoning’, Political Psychology, 35, 129–156. Leonisio, Rafael, Fernando Molina, and Diego Muro (2016). ETA’s Terrorist Campaign: From Violence to Politics, 1968–2015. Routledge. Lewis-Beck, Michael S (1990). Economics and elections: The major Western democracies. University of Michigan Press. Lewis-Beck, Michael S., and Richard Nadeau (2012). ‘PIGS or not? Economic voting in Southern Europe’, Electoral Studies, 31:3, 472–477. Lijphart, Arend (1979). ‘Religious vs. Linguistic vs. Class Voting: The “Crucial Experiment” of Comparing Belgium, Canada, South Africa, and Switzerland’, American Political Science Review, 73:02, 442–458. Linz, Juan J, SN Eisenstadt, and Stein Rokkan (1973). ‘Early state-building and late peripheral nationalism against the state: the case of Spain’,. Linz, Juan J., and Alfred Stepan (1996). Problems of Democratic Transition and Consolidation: Southern Europe, South America, and Post-Communist Europe. JHU Press. Lipset, Seymour M, and Stein Rokkan (1967a). ‘Cleavage Structures, Party Systems, and Voter Alignments: An Introduction’, in Seymour M Lipset and Stein Rokkan (eds.), Party Systems and Voter Alignments, vol. New York - London: The Free Press - Collier - Macmillan, 1–64. Lipset, S.M., and S. Rokkan (1967b). Party systems and voter alignments: cross-national perspectives. New York - London: The Free Press - Collier - Macmillan. Lubbers, Marcel, Mérove Gijsberts, and Peer Scheepers (2002). ‘Extreme right-wing voting in Western Europe’, European Journal of Political Research, 41:3, 345–378.

- 177 -

Lucardie, Paul (2000). ‘Prophets, Purifiers and Prolocutors Towards a Theory on the Emergence of New Parties’, Party politics, 6:2, 175–185. Lupu, Noam (2014). ‘Brand Dilution and the Breakdown of Political Parties in Latin America’, World Politics, 66:4, 561–602. Lupu, Noam (2016). Party brands in crisis: partisanship, brand dilution, and the breakdown of political parties in Latin America. New York, NY: Cambridge University Press. Lyrintzis, Christos (1984). ‘Political parties in post-junta Greece: A case of “bureaucratic clientelism”?’, West European Politics, 7:2, 99–118. Magalhães, Pedro C. (2005). ‘Disaffected democrats: Political attitudes and political action in Portugal’, West European Politics, 28:5, 973–991. Mainwaring, Scott (2006). ‘The Crisis of Representation in the Andes’, Journal of Democracy, 17:3, 13–27. Mair, Peter (1997a). Party system change: approaches and interpretations. Oxford : New York: Clarendon Press ; Oxford University Press. Mair, Peter (1997b). ‘E. E. Schattschneider’s The Semisovereign People’, Political Studies, 45:5, 947–954. Mair, Peter (2009). ‘Representative versus Responsible Government’, , Cologne: Max Planck Institute for the Study of Societies http://cadmus.eui.eu/handle/1814/12533 (Accessed April 7, 2015). Mair, Peter (2013). Ruling the void: the hollowing of Western democracy. Verso. Marks, Gary, Liesbet Hooghe, Moira Nelson, and Erica Edwards (2006). ‘Party Competition and European Integration in the East and West: Different Structure, Same Causality’, Comparative Political Studies, 39:2, 155–175. Martin, Pierre (2000). Comprendre les évolutions électorales: la théorie des réalignements revisitée. Les Presses de Sciences Po. Martínez, Ismael Crespo, and Alberto Mora Rodríguez (2015). ‘Los efectos del liderazgo en el comportamiento electoral en las elecciones de 2011 en España. La influencia de los atributos/The effects of leadership in voting behavior in elections of 2011, in Spain. The influence of the attributes’, Elecciones, 14:15, 145–170. Matthijs, Matthias (2014). ‘Mediterranean Blues: The Crisis in Southern Europe’, Journal of Democracy, 25:1, 101–115. Matthijs, Matthias (2017). ‘Integration at What Price? The Erosion of National Democracy in the Euro Periphery’, Government and Opposition, 52:2, 266–294. Matthijs, Matthias, and Mark Blyth, eds. (2015). The future of the euro. New York: Oxford University Press. Meguid, Bonnie M. (2005). ‘Competition between unequals: The role of mainstream party strategy in niche party success’, American Political Science Review, 99:03, 347–359. Mendez, Monica (2007). Turning the Page: Crisis and Transformation of the Spanish Socialist Party. London: Routledge. Michavila Nuñez, Narciso (2015). ‘El PP en la oposición’, in Eva Anduiza,Agusti Bosch,Lluis Orriols,and Guillem Rico (eds.), Elecciones generales 2011, vol., Elecciones. Madrid: Centro de Investigaciones Sociologicas, 47–62. Molinas, César ‘Theory of Spain’s political class’. El País, September 12 http://elpais.com/elpais/2012/09/12/inenglish/1347449744_053124.html (Accessed October 21, 2015). Molinas, César (2013). Qué hacer con España: del capitalismo castizo a la refundación de un país. Madrid: Ediciones Destino. Montalvo, Jose G. (2012). ‘Re-examining the evidence on the electoral impact of terrorist attacks: The Spanish election of 2004’, Electoral Studies, 31:1, 96–106.

- 178 -

Montero, José R., Francisco J. Llera, and Mariano Torcal (1992). ‘Sistemas Electorales en España: Una Recapitulación’, Reis, 58:58, 7–56. Montero, Jose Ramon (1998). ‘Sobre el sistema electoral español: rendimientos políticos y criterios de reforma’, in Juan Montabes (ed.), El sistema electoral a debate. Veinte años de rendimientos del sistema electoral a debate, vol. Madrid: Centro de Investigaciones Sociológicas / Parlamento Andaluz, 37–70. Moreno, Alejandro (1999). Political cleavages. Issues, parties, and the consolidation of Democracy. Boulder (Colorado): Westview Press. Morgan, Jana (2011). Bankrupt representation and party system collapse. Penn State Press. Müller-Rommel, Ferdinand (1998). ‘The new challengers: greens and right-wing populist parties in western Europe’, European Review, 6:02, 191–202. Muñoz, Jordi, and Raül Tormos (2015). ‘Economic expectations and support for secession in Catalonia: between causality and rationalization’, European Political Science Review, 7:02, 315–341. Muro, Diego (2009). ‘Territorial Accommodation, Party Politics, and Statute Reform in Spain’, South European Society and Politics, 14:4, 453–468. Muro, Diego (2018). ‘The Stillbirth of the Catalan Republic’, Current History, 117:797, 83. Muro, Diego, and Guillem Vidal (2016). ‘Political mistrust in southern Europe since the Great Recession’, Mediterranean Politics, 22:2, 197–217. Muro, Diego, and Martijn C. Vlaskamp (2016). ‘How do prospects of EU membership influence support for secession? A survey experiment in Catalonia and Scotland’, West European Politics, 39:6, 1115–1138. Newman, Saul (1997). ‘Ideological trends among ethnoregional parties in post‐industrial democracies’, Nationalism and Ethnic Politics, 3:1, 28–60. Noelke, Clemens (2016). ‘Employment Protection Legislation and the Youth Labour Market’, European Sociological Review, 32:4, 471–485. Oakley, David ‘Greece downgraded over high debt’. Financial Times, December 8 http://www.ft.com/intl/cms/s/0/2763a1d6-e3fc-11de-b2a9- 00144feab49a.html#axzz3pHeBhojo (Accessed October 22, 2015). Oesch, D. (2006a). ‘Coming to Grips with a Changing Class Structure: An Analysis of Employment Stratification in Britain, Germany, Sweden and Switzerland’, International Sociology, 21:2, 263–288. Oesch, Daniel (2006b). Redrawing the Class Map. http://www.palgrave.com/us/book/9781403985910 (Accessed December 3, 2016). Oesch, Daniel (2012). ‘The class basis of the cleavage between the New Left and the radical right’, Class Politics and the Radical Right,, 31. Oesch, Daniel, and Line Rennwald (2017). ‘Electoral competition in Europe’s new tripolar political space: class voting for the left, centre-right and radical right’,. Orriols, Lluis, and Guillermo Cordero (2016). ‘The Breakdown of the Spanish Two-Party System: The Upsurge of Podemos and Ciudadanos in the 2015 General Election’, South European Society and Politics, 21:4, 469–492. Orriols, Lluis, and Toni Rodon (2016). ‘The 2015 Catalan Election: The Independence Bid at the Polls’, South European Society and Politics, 21:3, 359–381. Pallarés, Francesc, and Michael Keating (2003). ‘Multi-Level Electoral Competition Regional Elections and Party Systems in Spain’, European Urban and Regional Studies, 10:3, 239–255. Pappas, Takis S., and Eoin O’Malley (2014). ‘Civil Compliance and “Political Luddism” Explaining Variance in Social Unrest During Crisis in Ireland and Greece’, American Behavioral Scientist, 58:12, 1592–1613.

- 179 -

Persson, Mikael (2015). ‘Education and Political Participation’, British Journal of Political Science, 45:3, 689–703. Polanyi, Karl (2001). The great transformation: The political and economic origins of our time. 2 Beacon Paperback. Boston, MA: Beacon Press. Polk, Jonathan, and Jan Rovny (2015). ‘Worlds of Welfare and Multidimensional Party Competition in Europe’, European University Institute, Florence. della Porta, Donatella (2015). Social Movements in Times of Austerity: Bringing Capitalism Back Into Protest Analysis. Cambridge, UK; Malden, MA: John Wiley & Sons. Powell, G. Bingham, Jr., and Guy D. Whitten (1993). ‘A Cross-National Analysis of Economic Voting: Taking Account of the Political Context’, American Journal of Political Science, 37:2, 391–414. Przeworski, Adam, and Henry Teune (1970). ‘The logic of comparative social inquiry’,. Ramiro, Luis, and Raul Gomez (2017). ‘Radical-left populism during the great recession: Podemos and its competition with the established radical left’, Political Studies, 65:1_suppl, 108–126. Rennwald, Line, and Geoffrey Evans (2014). ‘When Supply Creates Demand: Social Democratic Party Strategies and the Evolution of Class Voting’, West European Politics, 37:5, 1108–1135. Rhodes, Martin, ed. (1997). Southern European welfare states: between crisis and reform. London ; Portland, OR: Frank Cass. Rico, Guillem (2007). ‘“¡No nos falles!” Los candidatos y su peso electoral’, in Jose Ramon Montero,Ignacio Lago,and Mariano Torcal (eds.), Elecciones Generales 2004, vol., Elecciones. Madrid: Centro de Investigaciones Sociologicas. Rico, Guillem (2012). ‘The 2010 Regional Election in Catalonia: A Multilevel Account in an Age of Economic Crisis’, South European Society and Politics, 17:2, 217–238. Rico, Guillem, and M. Kent Jennings (2012). ‘The Intergenerational Transmission of Contending Place Identities: Intergenerational Transmission of Contending Place Identities’, Political Psychology, 33:5, 723–742. Rico, Guillem, and Robert Liñeira (2014). ‘Bringing Secessionism into the Mainstream: The 2012 Regional Election in Catalonia’, South European Society and Politics, 19:2, 257–280. Riera, Pedro, Pablo Barberá, Raúl Gómez, Juan Antonio Mayoral, and José Ramón Montero (2013). ‘The electoral consequences of corruption scandals in Spain’, Crime, Law and Social Change, 60:5, 515–534. Roberts, Kenneth (2013). ‘Market Reform, Programmatic (De)alignment, and Party System Stability in Latin America’, Comparative Political Studies, 46:11, 1422–1452. Roberts, Kenneth M. (2017). ‘State of the Field: Party politics in hard times: Comparative perspectives on the European and Latin American economic crises’, European Journal of Political Research, http://doi.wiley.com/10.1111/1475-6765.12191 (Accessed January 30, 2017). Rodríguez Teruel, Juan, and Astrid Barrio (2016). ‘Going National: Ciudadanos from Catalonia to Spain’, South European Society and Politics, 21:4, 587–607. Rodríguez-Teruel, Juan, Astrid Barrio, and Oscar Barberà (2016). ‘Fast and Furious: Podemos’ Quest for Power in Multi-level Spain’, South European Society and Politics, 21:4, 561–585. Rodrik, Dani (2013). ‘The Globalization Paradox’, Democracy and the Future ofthe World Economy,. Rohrschneider, Robert (1993). ‘New Party versus Old Left Realignments: Environmental Attitudes, Party Policies, and Partisan Affiliations in Four West European Countries’, The Journal of Politics, 55:03, 682.

- 180 -

Rohrschneider, Robert, and Stephen Whitefield (2012). The Strain of Representation: How Parties Represent Diverse Voters in Western and Eastern Europe. Oxford, New York: Oxford University Press. Rokkan, Stein (1999). State Formation, Nation-building, and Mass Politics in Europe: The Theory of Stein Rokkan : Based on His Collected Works. Oxford University Press. Romero, Joan (2012). ‘España inacabada. Organización territorial del Estado, autonomía política y reconocimiento de la diversidad nacional’, Documents d’anàlisi geogràfica, 58:1, 13–49. Roth, Felix, Felicitas Nowak-Lehmann D., and Thomas Otter (2013). ‘Crisis and trust in national and European Union institutions: panel evidence for the EU, 1999 to 2012’, http://cadmus.eui.eu/handle/1814/26975 (Accessed July 16, 2013). Rovny, Jan, and Jonathan Polk (2014). ‘New Wine in Old Bottles: Explaining Party Competition Along the Socio-Cultural Dimension in Europe?’, in Workshop on party competition and voter alignments in times of welfare state transformations, vol. Sciences Po, Paris, France http://jonathanpolk.net/wp- content/uploads/2013/08/The_Other_Dim_Nov5_2014JR.pdf (Accessed May 2, 2016). Royo, Sebastián (2014). ‘Institutional Degeneration and the Economic Crisis in Spain’, American Behavioral Scientist, 58:12, 1568–1591. Rueda, David (2007). Social democracy inside out: partisanship and labor market policy in industrialized democracies. Oxford ; New York: Oxford University Press. Ruiz Jimenez, Antonia Maria (2007). ‘Competición política y representación democrática: La oferta electoral de los partidos’, in Jose Ramon Montero,Ignacio Lago,and Mariano Torcal (eds.), Elecciones generales 2004, vol. Madrid: Centro de Investigaciones Sociologicas, 65–106. Ruiz Jimenez, Antonia Maria (2010). ‘Los temas de la campaña electoral: ¿Electorado diverso, temas divergentes?’, in Jose Ramon Montero and Ignacio Lago (eds.), Elecciones Generales 2008, vol., Elecciones. Madrid: Centro de Investigaciones Sociologicas, 93–122. Sampedro, Victor, and Francisco Seoane Perez (2009). ‘Las elecciones generales españolas de 2008: “bipolarización antagónica” fomentada por intereses político-mediáticos y las nuevas tecnologías’, Revista de Sociologia e Política, 17:34, 129–135. Sánchez de Dios, Manuel (1999). ‘Parliamentary Party Discipline in Spain’, in Shaun Bowler,David M Farrell,and Richard S Katz (eds.), Party discipline and Parliamentary Government, vol. Columbus: Ohio State University Press, 141–162. Sartori, Giovanni (1969). ‘From the sociology of politics to political sociology’, Government and Opposition, 4:2, 195–214. Scharpf, Fritz W (1999). Governing in Europe: effective and democratic? Oxford University Press. Scharpf, Fritz W. (2011). ‘Monetary Union, Fiscal Crisis and the Pre-emption of Democracy’, Zeitschrift für Staats- und Europawissenschaften (ZSE) / Journal for Comparative Government and European Policy, 9:2, 163–198. Schattschneider, Elmer Eric (1960). The semisovereign people: A realist’s view of democracy in America. New York: Holt, Rinehart and Winston. Schimmelfennig, Frank (2014). ‘European Integration in the Euro Crisis: The Limits of Postfunctionalism’, Journal of European Integration, 36:3, 321–337. Serrano, Ivan (2017). ‘Normative views in practice: the political discourse on secession in Catalonia 2003-2015’, , 21. Serrano, Ivan, and Albert Bonillo (2017). ‘Boundary shifts and vote alignment in Catalonia’, Ethnicities, 17:3, 371–391.

- 181 -

Sides, John (2006). ‘The Origins of Campaign Agendas’, British Journal of Political Science, 36:03, 407. Simón, Pablo (2017). ‘The Challenges of the New Spanish Multipartism: Government Formation Failure and the 2016 General Election’, South European Society and Politics, 21:4, 493–517. Statham, Paul, and Hans-Jorg Trenz (2015). ‘Understanding the mechanisms of EU politicization: Lessons from the Eurozone crisis’, Comparative European Politics, 13:3, 287–306. Stavrakakis, Yannis, and Giorgos Katsambekis (2014). ‘Left-wing populism in the European periphery: the case of SYRIZA’, Journal of Political Ideologies, 19:2, 119–142. Steenbergen, Marco R., Erica E. Edwards, and Catherine E. de Vries (2007). ‘Who’s Cueing Whom?: Mass-Elite Linkages and the Future of European Integration’, European Union Politics, 8:1, 13–35. Steiner, Nils D., and Christian W. Martin (2012). ‘Economic Integration, Party Polarisation and Electoral Turnout’, West European Politics, 35:2, 238–265. Taylor, Michael, and V. M. Herman (1971). ‘Party Systems and Government Stability’, The American Political Science Review, 65:1, 28–37. Teruel, Juan Rodríguez, and Astrid Barrio (2015). ‘Going National: Ciudadanos from Catalonia to Spain’, South European Society and Politics, 0:0, 1–21. Torcal, Mariano (2010). ‘Los anclajes del voto en las elecciones de 2008’, in Jose Ramon Montero and Ignacio Lago (eds.), Elecciones Generales 2008, vol., Elecciones. Madrid: Centro de Investigaciones Sociologicas, 269–302. Torcal, Mariano (2014). ‘The Decline of Political Trust in Spain and Portugal Economic Performance or Political Responsiveness?’, American Behavioral Scientist, 58:12, 1542–1567. Torcal, Mariano, and Lucía Medina (2007). ‘La competencia electoral entre PSOE y PP: el peso de los anclajes de ideología, religión y clase’, in Jose Ramon Montero,Ignacio Lago,and Mariano Torcal (eds.), Elecciones Generales 2004, vol., Elecciones. Madrid: Centro de Investigaciones Sociologicas, 275–302. Tremlett, Giles ‘Spain changes constitution to cap budget deficit’. The Guardian, August 26 http://www.theguardian.com/business/2011/aug/26/spain-constitutional-cap-deficit (Accessed October 22, 2015). Tremlett, Giles ‘Mariano Rajoy announces €65bn in austerity measures for Spain’. The Guardian, July 11 http://www.theguardian.com/business/2012/jul/11/mariano-rajoy- spain-65bn-cuts (Accessed October 19, 2015). Tsatsanis, Emmanouil, André Freire, and Yannis Tsirbas (2014). ‘The Impact of the Economic Crisis on the Ideological Space in Portugal and Greece: A Comparison of Elites and Voters’, South European Society and Politics, 19:4, 519–540. Urdánoz, Jorge (2008). ‘Ingeniería del sufragio. Artimañas electorales vs. Valores democráticos’, Claves de Razón Práctica, 180, 58–67. Urquizu, Ignacio (2015). ‘La Segunda Legislatura de Rodríguez Zapatero’, in Eva Anduiza,Agusti Bosch,Lluis Orriols,and Guillem Rico (eds.), Elecciones generales 2011, vol., Elecciones. Madrid: Centro de Investigaciones Sociologicas, 29–46. Verney, Susannah (2014). ‘“Broken and Can’t be Fixed”: The Impact of the Economic Crisis on the Greek Party System’, The International Spectator, 49:1, 18–35. Vidal, Guillem (2018). ‘Challenging business as usual? The rise of new parties in Spain in times of crisis’, West European Politics, 41:2, 261–286. Villodres, Carmen Ortega, and Juan Montabes Pereira (2002). ‘El voto limitado en las elecciones al Senado español: estrategias de nominación y rendimientos partidistas en las elecciones de marzo de 2000’, Revista española de ciencia política,7, 103–130.

- 182 -

Wolf, Martin ‘Merkozy failed to save the eurozone’. Financial Times, December 6 http://www.ft.com/intl/cms/s/0/396ff020-1ffd-11e1-8662- 00144feabdc0.html#axzz3p05Sd600 (Accessed October 19, 2015). Wright, Erik Olin (2015). Understanding class. Verso Books. Zaller, John (2012). ‘What Nature and Origins Leaves Out’, Critical Review, 24:4, 569–642. Zaller, John R (1992). The nature and origins of mass opinion. Cambridge university press. Zapata-Barrero, Ricard (2009). ‘Policies and public opinion towards immigrants: the Spanish case.’, Ethnic & Racial Studies, 32:7, 1101–1120. Zechmeister, Elizabeth J., and Daniel Zizumbo-Colunga (2013). ‘The Varying Political Toll of Concerns About Corruption in Good Versus Bad Economic Times’, Comparative Political Studies, 46:10, 1190–1218.

- 183 -

- 184 -

Appendices

Appendix Ch. 3: Old versus new politics: The political spaces in Southern Europe in times of crisis

Data sources and coding strategy

While media data come with certain biases, we think they offer ample opportunities to capture changes in the political space in times of crises. More precisely, we rely on media data because we are interested in publicly visible conflicts among the parties during the campaigns. In our opinion, media data are especially sensitive to political change and allow us to examine how the issues of the day map onto underlying issue dimensions. While this might lead to limited information about small parties (as they might be underreported in the media), it gives a good indication of the conflicts and actors that dominate the public debate.

Alternative data sources do not come with the same biases. However, they are usually not linked to specific elections (especially expert surveys), do not contain positional and salience measure for all issues (especially manifesto data), and apply a rather rigid issue set of issue categories (which we tend to avoid by relying on a more inductive approach to new issues).

As stated, we selected articles from two quality newspapers per country (the main center-left and center-right newspapers): Ta Nea and Kathimerini (Greece), La Repubblica and Corriere della Sera (Italy), Público and Diário de Notícias (Portugal), and El Pais and El

Mundo (Spain). We selected all news articles that were published within two months before the national Election Day and reported on the electoral contest and national party politics more generally. Editorials and commentaries were excluded from the selection. The selection was done by an extensive keyword list including the names and abbreviations of political parties and key politicians from each party. In the case of early elections, we selected the period from the announcement of the election until Election Day.

- 185 -

We then coded a sample of the selected articles using core sentence analysis (CSA).

Following this type of relational content analysis, each grammatical sentence of an article is reduced to its most basic ‘core sentence(s)’ structure, which contain(s) only the subject, the object, and the direction of the relationship between the two. The core sentence approach was developed by Kleinnijenhuis and colleagues (e.g., Kleinnijenhuis, et al., 1997). This type of quantitative content analysis allows us to study both issue positions and salience. The direction between actors and issues is quantified using a scale ranging from -1 to +1, with three intermediary positions. For example, the grammatical sentence “Party leader A rejects calls for leaving the Eurozone but supports a haircut on the country’s debt” leads to two coded observations (Party A +1 Eurozone membership; Party A +1 haircut). For this paper, we only focus on relations between party actors and political issues, that is we neglect relations between different actors. The following analysis are based on a total of around

11,000 actor-issue statements, ranging from a minimum of 942 for the first Greek election in

2012 to a maximum of 1,650 for the Portuguese election in 2011 (see Table 1A).

Table 1A. Number of actor-issue sentences per campaign Country Election Number of sentences Greece 2012-May 942 Greece 2012-June 1,027 Greece 2015-January 1,058 Greece 2015-September 975 Italy 2013 1,521 Portugal 2011 1,650 Portugal 2015 1,345 Spain 2011 1,151 Spain 2015 1,406 Total 11,075 11,075

Note: The table shows the total number of actor-issue statements used for the analysis by campaign.

We coded the function, party affiliation, and (if available) name of actors. For the present analysis, the actors were grouped according to their party affiliation. The issues were coded in even more detail (with more than 200 coded categories per election campaign). As described in the main text, the issues were aggregated into 18 categories that allow us to (a)

- 186 -

capture the broad conflicts described in the previous sections and (b) compare our results to previous results of Kriesi et al. (2008, 2012)(2008, 2012). Descriptions of the 18 categories can be found in Table 2A. Furthermore, we indicate what a positive position regarding a given issue category indicates.

Table 2A. Issue categories

Categories Description (a position of +1 stands for ...) Domestic economic 1. welfare support for an expansion of the welfare state; objection to welfare state retrenchment; support for tax reforms with a redistributive character; calls for employment and health care programs 2. economic liberalism opposition to market regulation, economic protectionism in agriculture and other sectors of the economy; support for deregulation, more competition, and privatization; support for a rigid budgetary policy; reduction of the state deficit and taxes without direct redistributive effects 3. economic reform (vague) support for general economic reforms without a clear direction (e.g., fighting economic crisis; fighting unemployment) Domestic political 4. democratic renewal support for institutional reforms to make the political system more democratic or transparent; opposition to corruption and political class 5. democratic reform (vague) support for general reforms of the political system without clear direction 6. regionalism support for regional autonomy or independence 7. media diversity support for equal media access and coverage of parties European economic 8. euro support for the single European currency; opposition to a country leaving the Eurozone 9. anti-bailout opposition to the bailout 10. anti-bailout (conditions) opposition to the conditions of the bailout; support for more favorable conditions (e.g., better interest rates, debt restructuring) European political 11. europe support for European integration in general, deepening and widening Others 12. cultural liberalism support for cultural diversity, international cooperation, gender equality, LGTB rights; opposition to national traditions and traditional moral values. 13. anti-immigration support for restrictive immigration and integration policies 14. security support for more law and order, fighting crime (except tax fraud and corruption) 15. defense support for military interventions, the armed forces, a strong national defense, and nuclear weapons 16. education support for education and research

- 187 -

17. environment support for environmental protection; opposition to nuclear energy 18. infrastructure support for improving the country’s roads, railways, and other physical infrastructure; support for media

Salience and polarization

The systemic components of party competition that we are most interested in are operationalized by the salience and the polarization of the issues which represent the economic and political conflicts. Salience is measured by the share of core sentences on an issue category in percent of all sentences related to any issue. The indicator for the polarization of party positions is based on Taylor and Hermann’s (1971) index and ranges from 0 to 1 (=maximum degree of polarization). The polarization of positions on a given issue category is computed as follow:

K POLARIZATION = w x - x 2 å k ( k ) k=1

Where !" is the salience of a particular issue category for party k, #" is the position of party k on this issue category, and #$ is the weighted average position of all parties, where weights are provided by the party-specific salience of the issue category.

Table 3A. Salience of economic and political issues by election and country

Economic Political Others

Domestic European Domestic European GR 2012-May 34.3 23.7 22.8 6.9 12.3 GR 2012-June 27.3 36.5 9.0 9.5 17.7 GR 2015-Jan 30.8 32.0 22.2 7.9 7.0 GR 2015-Sept 29.1 27.9 29.3 4.7 8.9 IT 2013 50.0 1.2 30.6 3.4 14.8 PT 2011 53.7 14.1 18.2 0.3 13.6 PT 2015 64.1 2.9 10.1 1.7 21.2 ES 2011 58.6 1.9 18.1 1.1 20.3 ES 2015 26.6 1.0 52.2 0.3 20.0

Note: The Table shows the share of an issue category in percent of all coded issues by campaign.

- 188 -

Multi-dimensional scaling: procedure and interpretation

To analyze the structure of the political space and the resulting actor configuration, we combine the visibility of the actors, their issue positions and issue emphasis. To do so, we follow Kriesi et al. (2006, 2008) and rely on multi-dimensional scaling (MDS). For this analysis, we excluded issue categories with less than three percent of all statements and actors with less than 20 statements. In the case of Spain, we also dropped all regional parties to concentrate on party competition at the national level.

The aim of MDS is to obtain a graphical representation of the relative locations of a set of objects in a low-dimensional space. Starting from information on the proximities between pairs of objects, MDS finds the optimal space in which these objects can be located while distorting as little as possible the original proximities. In our case, we consider the distances between parties and issues, which were determined with the help of content analyses, and use MDS to construct a representation of the locations of these parties and issues in a (two-dimensional) space. It is not our purpose here to explain all details of MDS.

Overviews of this statistical method can be found in Borg and Groenen (1997), Cox and Cox

(2001), or Kruskal and Wish (1978). Rabinowitz (1975) also presents a good and simple introduction to the topic but limited to the case of nonmetric MDS. Rather, we simply want here to emphasize some of the peculiarities of the analyses we have performed, which differ from a ‘standard’ MDS. Most important among these are the unfolding technique and the use of weights.

Unfolding models are special cases of MDS where information is available on the proximities between objects of two different sets, but not on the proximities between objects within each of the two sets. In our analyses, we used information on the distances between parties and issues. But we had no comparable measures of the distances between pairs of parties or between pairs of issues. Such data can be analyzed with standard MDS techniques,

- 189 -

but it means that at least half of the cells in the proximity matrix have missing values. When such data are analyzed with nonmetric MDS (which implies a transformation of the original proximities into rank orders, for instance), the solutions ‘are subject to many potential degeneracies’ (Borg and Groenen, 1997: 231). Metric MDS, however, is more robust and avoids the problems linked with nonmetric MDS in the case of unfolding (Borg and Groenen,

1997: 245ff.).

The second particularity of our MDS analyses is the use of weights. As explained above, MDS locates the objects in a space while keeping the distances between them as close as possible to the original proximities. However, as the aim is to obtain a representation of the objects in a low dimensional space, some distortion of the original distances is unavoidable. The degree of distortion is measured by a ‘Stress’ statistic, which is based on the sum of the squared distances between the original proximities and the proximities obtained in the solution. The higher the value of the Stress statistic, the worse is the fit between the solution and the data. The aim is thus to find the solution that minimizes the value of Stress. In our case, we want to obtain a configuration of parties and issues where the distances between them are as close as possible to those in our original data. By using weights, we allow for the possibility that some distortions of original distances have a larger impact on the value of Stress than others. In other words, we give much importance to representing some distances faithfully and less importance to the degree of distortion affecting other distances. The rationale for this is that not all relations between parties and issues have the same importance. In a campaign, parties address some issues very frequently while other issues play a minor role in their statements. Similarly, not all parties are central actors in a campaign. We account for such variation by computing weights that reflect the salience of a given party × issue relationship. These weights are calculated as the number of

- 190 -

core sentences corresponding to a given party × relationship, expressed as a proportion of the total number of relationships between parties and issues.

As in any spatial analysis, it is important to note that the focus on dimensions comes at the expense of details. That is, MDS allows identifying main lines of opposition in the party system like under a magnifying glass. The tradeoff is that less salient issues and actors are less accurately represented. MDS configurations can only be interpreted regarding distances between objects. The orientation of a configuration is arbitrary, which implies that it can be freely rotated. To facilitate comparison of the spaces shown in the manuscript, we have rotated them in such a way that the issues ‘welfare’ and ‘economic liberalism’ are situated on a horizontal line. To focus attention on those aspects of these configurations that are most important given our hypotheses, we have added four graphical elements. First, we have drawn a line connecting welfare and economic liberalism as the supposed endpoints of the economic left-right divide. Second, we have introduced circles of varying size around the points identifying issues and parties to indicate their respective salience. Third, we added a line connecting ‘anti-bailout’ and ‘pro-euro’ for the Greek elections, and a vertical dashed line at the mid-point of the left-right divide for all other elections. Finally, we have added a circle that covers the main left-wing challengers and the issue of ‘democratic renewal.' This circle is added to facilitate the interpretation of our results regarding the alignment of opposition to ‘old politics’ and conflicts over austerity.

Moreover, to systematically compare the alignment of conflicts over economic and political issues, we calculate a measure for the integration of ‘democratic renewal’ in the left- wing pole of the left-right dimension. To do so, we divided the distance between ‘democratic renewal’ and ‘welfare’ (the left-wing pole of the economic dimension) by the distance between ‘democratic renewal’ and ‘economic liberalism’ (the right-wing pole of the dimension). We rely on such a relative measure because the distances in a MDS graph can

- 191 -

only be interpreted relative to each other. Moreover, to systematically compare situations in which ‘democratic renewal’ is more integrated in the left-wing pole with those where it is more integrated in the right-wing pole, we present the natural log of the division in Table 4A.

Overall, the values in table 4A highlight the left-wing integration of calls for democratic renewal. The only exception is the Italian campaign in 2013 as indicated by the slightly negative value of -0.5. Moreover, the comparisons over time within a given country, support our third hypothesis as the left-wing integration is stronger in cases when the mainstream left has been in opposition (the most telling example is Portugal).

Table 4A. Integration of ‘democratic renewal’ in left-wing pole

Distance of democratic renewal to ln of

welfare eco-liberalism y/x left pole (x) right pole (y) GR 2012-May 1.0 1.2 0.2 GR 2012-June 1.0 1.5 0.4 GR 2015-Jan 1.0 3.4 1.2 GR 2015-Sept 1.0 2.8 1.0 IT 2013 1.0 0.6 -0.5 PT 2011 1.0 1.4 0.4 PT 2015 1.0 9.0 2.2 ES 2011 1.0 2.7 1.0 ES 2015 1.0 5.7 1.7

Note: The table shows distances in the political spaces presented in Figure 2. More precisely, it shows the distances between the issue ‘democratic renewal’, on the one side, and ‘welfare (left pole)’ respectively ‘economic liberalism (right pole)’, on the other. As the absolute measures make little sense, we standardized them so that the distance between democratic renewal and welfare equals 1. The last column shows the key indicator, i.e., the logged value of the division. It takes a value of zero if democratic renewal is equally distant from both poles, while it gets positive, the more integrated it is in the left-wing pole.

References Borg, Ingwer and Patrick Groenen (1997). Modern Multidimensional Scaling: Theory and Applications. Springer Series in Statistics. New York: Springer. Cox, Trevor F. and Michael A. A. Cox (2001). Multidimensional Scaling. Second Edition. London: Chapman & Hall.

- 192 -

Kruskal, Joseph B. and Myron Wish (1978). Multidimensional Scaling. Sage University Paper series on Quantitative Application in the Social Sciences, 07-011. Newbury Park: Sage. Rabinowitz, George B. (1975). “An Introduction to Nonmetric Multidimensional Scaling”, American Journal of Political Science 19(2): 343-390. Taylor, Michael and V.M. Herman (1971). “Party Systems and Government Stability.” American Political Science Review 65(1): 28-37.

- 193 -

Appendix Ch. 4: Out with the Old: Restructuring Spanish Politics

Table 1A. Overview of events and elections by legislature

Legislature Government Main events during the legislature Decentralization of competences, including health care management Jose Maria Aznar 2000-2004 Participation in the Iraq war as part of the Açores Pact (PP) 11March Al-Qaeda terrorist attacks Government pulls troops out of Iraq Advancement in civil rights (abortion, dependence, equal marriage) leading to high levels of conflict with Catholic Church Negotiations to surrender ETA Jose Luis Rodriguez Basque government proposes a renegotiation of autonomy (refused 2004-2008 Zapatero (PSOE) in Congress) Passing of the Catalan Autonomy Act PP encourages high levels of polarization in public debate (amongst others, they blamed Rodriguez Zapatero for the 11-March terrorist attacks as a strategy to steal the election) Economic situation can be divided into two periods: • First half of the legislature: implementation of previous reforms • Second half of the legislature: plans to stimulate consumption (plan E, allocations to families) + refusal to recognize the Jose Luis Rodriguez economic crisis 2008 - 2011 Zapatero (PSOE) New party: UPyD ETA declares the ceasing of terrorist actions Constitutional Court declares unconstitutional the Catalan Autonomy Act 15-M demonstrations (May 2011) Retrenchment and austerity policies Banking system bailout Corruption scandals Restriction of civil liberties (gag law) Mariano Rajoy Brey 2011 - 2015 Intense period of demonstrations building on some structures formed (PP) after the 15M mobilizations (“mareas”) Situation with Catalonian government reaches a high point of tension after attempt to launch an independence process New parties: Podemos & Ciudadanos Negotiations to form a coalition government fail 2015-2016 - Corruption scandals affecting mainly PP keep appearing New elections held on June 26, 2016

- 194 -

Table 2A. Spain 2004-2015: brief overview of the elections

Election Incumbent government Outcome government Majority government Minority government 2004 PP PSOE Minority government Minority government 2008 PSOE PSOE Minority government Majority government 2011 PSOE PP

Majority government No party is able to form 2015 PP government, new elections are held Minority Government 2016 No government PP

- 195 -

Appendix Ch. 5: Challenging business as usual? The rise of new parties in Spain in times of crisis

Table A.0 Descriptive Statistics Variables N Mean SD Min Max Categories

Age 12,417 50.01 17.81 18 98 Gender 12,417 0.517 0.5 0 1 0 = Male; 1= Female Size Town/City 12,417 3.774 1.664 1 7 1 <= 2000; 2 = 2-10k; 3 = 10-50k; 4 = 50-100k; 5 = 100-400k; 6 = 400-1000K; 7 => 1000K Education 12,382 1.953 0.678 1 3 1 = Low; 2 = Middle; 3 = Higher Ideology (left-right) 10,547 4.626 2.004 1 10 1 = Left; 10 = Right Economic Liberalism 11,645 3.843 2.300 0 10 0 = More Public Services at the cost of more tax; 10 = Less tax at the cost of less public services Multiculturalism 12,085 3.855 2.638 0 10 0 = Nativism; 10 = Cosmopolitanism Civil Liberties 11,811 5.304 2.460 0 10 0 = Law and Order; 10 = Civil Liberties Territorial Organization 11,502 3.183 1.191 1 5 1 = Centralization; 5 = Decentralization Unemployed 12,367 0.161 0.367 0 1 1 = Unemployed Income 9,456 3.833 2.010 1 8 1 = None; 2 = <300; 3 = 300-600; 4 = 600-900; 5 = 900-1200; 6 = 1200-1800; 7 = 1800-2400; 8 = 2400+ Interest in Politics 12,380 2.357 0.967 1 4 1 = None; 2 = Little; 3 = Quite; 4 = A Lot Abstention 12,417 0.120 0.325 0 1 1 = Abstained in Previous Elections Sociotropic 12,358 0.478 0.234 0 1 [Normalized] 0 = Good/V.Good; 0.33 = Average; 0.66 = Bad; 1 = Very Bad. Weighted by Sociotropic Salience Sociotropic Salience * 12,417 4.468 1.691 1 7 1 = Neither mentioned; 2 = None and Third; (concern Economy or 3 = None and Second; 4 = Third and Second; Unemployment) 5 = First and None; 6 = Third and First; 7 = First and Second Pocketbook 12,367 0.304 0.264 0 1 [Normalized] 0 = Good/V.Good; 0.33 = Average; 0.66 = Bad; 1 = Very Bad Political Crisis 12,320 0.375 0.271 0 1 Based on Factor Analysis (see Table A.1) and weighted by Political Crisis Salience Political Crisis Salience 12,417 2.246 1.117 1 4 1 = Not mentioned; 2 = Third Problem; (concern of corruption) 3 = Second Problem; 4 = First Problem Region 12,417 8.045 4.800 1 19 A dummy for each autonomous community in Spain Social Class 11,964 4.991 2.195 1 8 See Figure A.1 for details.

Notes: * The question about the most important problem facing the country offers two answers on the economy: Unemployment and The Economy in General. Thus, salience considers both possible answers in hierarchical order (first, second and third most important problems), which renders 7 possibilities as they cannot both be selected together.

- 196 -

Table A.1 Factor analysis results

Variable Factor Uniqueness

Satisfaction with Democracy .77 0.41 Trust in Parliament .89 0.21 Trust in Political Parties .85 0.28

Eigenvalues 2.11 Variance 1.53 Note: Principal-Component Factors after varimax rotation. All variables range from 0 to 10, where 0 is full trust/satisfaction and 0 is no trust/satisfaction at all.

- 197 -

Table A.2 Marginal Effects of Different Operationalisations of Political Crisis Indicators on Vote (Robustness Check) PP PSOE Podemos Ciudadanos

Variables M1.A M2.A M3.A M1.B M2.B M3.B M1.C M2.C M3.C M1.D M2.D M3.D

Dissatisfaction -0.033*** -0.040*** 0.035*** 0.008**

w/ Democracy (0.004) (0.005) (0.004) (0.004)

Mistrust Parties -0.022*** -0.025*** 0.017*** 0.011***

(0.003) (0.004) (0.004) (0.003)

Mistrust Parliament -0.028*** -0.038*** 0.026*** 0.009**

(0.004) (0.004) (0.004) (0.004)

Observations 9,093 9,057 8,825 9,093 9,057 8,825 9,093 9,057 8,825 9,093 9,057 8,825

Notes: Standard errors in parentheses. Levels of statistical significance: *** p<0.01, ** p<0.05, * p<0.1. Each of the variables are weighted by “Corruption Salience”. All models are based on Model 2 in Table 3.

- 198 -

Table A.3 Pairwise correlations between ideological dimensions

Left-Right Public Services Multicul- Security Territorial Ideological Positions Scale vs Tax turalism & Freedom Organization

Left-Right Scale 1

Public Services vs Tax .16 1

Multiculturalism .3 .34 1

Security & Freedom .34 .21 .35 1

Territorial Organization .36 .1 .2 .27 1

Note: All correlations are significant at p<0.01.

Table A.3 Marginal Effects of Multinomial Logistic Regressions on Vote for Podemos

DV: Vote for Podemos Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7

Age -0.003*** -0.003*** -0.003*** -0.003*** -0.003*** -0.003*** -0.003*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Gender -0.038*** -0.042*** -0.040*** -0.026*** -0.026*** -0.026*** -0.026*** (0.008) (0.009) (0.009) (0.008) (0.009) (0.008) (0.008) Size Town/City 0.015*** 0.015*** 0.015*** 0.013*** 0.014*** 0.013*** 0.013*** (0.002) (0.003) (0.003) (0.002) (0.003) (0.002) (0.002) Education (Ref. Higher) Low -0.064*** -0.059*** -0.060*** -0.035** -0.032* -0.035** -0.034** (0.015) (0.018) (0.018) (0.015) (0.018) (0.015) (0.015) Middle -0.022** -0.020 -0.020 -0.008 -0.007 -0.008 -0.008 (0.011) (0.013) (0.013) (0.011) (0.012) (0.011) (0.011) Ideology -0.086*** -0.087*** -0.086*** -0.080*** -0.079*** -0.079*** -0.079*** (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) Unemployment 0.009 0.008 0.005 (0.012) (0.012) (0.012) Income 0.002 0.003 0.002 (0.003) (0.003) (0.003) Interest Politics 0.054*** 0.059*** 0.054*** 0.054*** (0.004) (0.004) (0.004) (0.004) Abstention 0.010 0.008 0.010 0.010 (0.012) (0.014) (0.012) (0.012) Economic Voting Sociotropic 0.040** 0.066*** (0.018) (0.019) Pocketbook 0.028* 0.034** (0.017) (0.017) Political Crisis 0.081*** 0.098*** 0.082*** 0.081*** (0.013) (0.016) (0.013) (0.013) Political Crisis X Ideology 0.086*** (0.019)

Political Crisis X Age 0.082*** (0.014)

Controls yes yes yes yes yes yes yes

199

Observations 9,251 7,335 7,350 9,202 7,280 9,202 9,202 Standard errors in parentheses (*** p<0.01, ** p<0.05, * p<0.1)

Table A.4 Marginal Effects of Multinomial Logistic Regressions on Vote for Ciudadanos

DV: Vote for Ciudadanos Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7

Age -0.002*** -0.002*** -0.002*** -0.002*** -0.002*** -0.002*** -0.002*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Gender -0.002 0.004 0.003 0.002 0.007 0.002 0.002 (0.006) (0.008) (0.008) (0.006) (0.008) (0.006) (0.006) Size Town/City 0.003 0.004* 0.005** 0.002 0.004* 0.002 0.002 (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) Education (Ref. Higher) Low -0.067*** -0.053*** -0.051*** -0.058*** -0.043*** -0.058*** -0.057*** (0.012) (0.015) (0.015) (0.012) (0.015) (0.012) (0.012) Middle -0.031*** -0.028** -0.027** -0.027*** -0.023** -0.027*** -0.027*** (0.010) (0.011) (0.011) (0.010) (0.011) (0.010) (0.010) Ideology 0.031*** 0.032*** 0.031*** 0.031*** 0.032*** 0.031*** 0.031*** (0.001) (0.002) (0.001) (0.001) (0.002) (0.001) (0.001) Unemployment 0.006 0.010 0.012 (0.010) (0.011) (0.011) Income 0.006*** 0.005** 0.005** (0.002) (0.002) (0.002) Interest Politics 0.016*** 0.015*** 0.016*** 0.016*** (0.003) (0.004) (0.003) (0.003) Abstention -0.001 0.000 -0.001 -0.001 (0.010) (0.012) (0.010) (0.010) Economic Voting Sociotropic -0.010 0.002 (0.015) (0.016) Pocketbook -0.033** -0.026* (0.015) (0.015) Political Crisis 0.030** 0.029** 0.030** 0.030** (0.012) (0.013) (0.012) (0.012) Political Crisis X Ideology 0.096*** (0.026)

Political Crisis X Age 0.031** (0.012)

Controls yes yes yes yes yes yes yes Observations 9,251 7,335 7,350 9,202 7,280 9,202 9,202 Standard errors in parentheses (*** p<0.01, ** p<0.05, * p<0.1)

200

Table A.5 Marginal Effects of Multinomial Logistic Regressions on Vote for PSOE

DV: Vote for PSOE Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7

Age 0.003*** 0.004*** 0.004*** 0.003*** 0.004*** 0.003*** 0.003*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Gender 0.044*** 0.045*** 0.042*** 0.036*** 0.034*** 0.036*** 0.035*** (0.008) (0.010) (0.010) (0.008) (0.010) (0.008) (0.008) Size Town/City -0.006** -0.004 -0.004 -0.005* -0.003 -0.005* -0.005* (0.003) (0.003) (0.003) (0.003) (0.003) (0.003) (0.003) Education (Ref. Higher) Low 0.099*** 0.093*** 0.097*** 0.097*** 0.094*** 0.097*** 0.097*** (0.017) (0.020) (0.020) (0.017) (0.020) (0.017) (0.017) Middle 0.050*** 0.048*** 0.050*** 0.047*** 0.047*** 0.047*** 0.047*** (0.012) (0.014) (0.014) (0.012) (0.014) (0.012) (0.012) Ideology -0.051*** -0.049*** -0.051*** -0.051*** -0.051*** -0.051*** -0.051*** (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) Unemployment -0.003 0.007 0.009 (0.013) (0.013) (0.013) Income 0.001 -0.001 -0.001 (0.003) (0.003) (0.003) Interest Politics -0.004 -0.009* -0.004 -0.004 (0.004) (0.005) (0.004) (0.004) Abstention -0.124*** -0.131*** -0.124*** -0.124*** (0.011) (0.013) (0.011) (0.011) Economic Voting Sociotropic 0.034* 0.027 (0.020) (0.020) Pocketbook -0.054*** -0.049*** (0.019) (0.019) Political Crisis -0.083*** -0.075*** -0.082*** -0.083*** (0.015) (0.017) (0.015) (0.015) Political Crisis X Ideology -0.06*** (0.023) Political Crisis X Age -0.09*** (0.015)

Controls yes yes yes yes yes yes yes Observations 9,251 7,335 7,350 9,202 7,280 9,202 9,202 Standard errors in parentheses (*** p<0.01, ** p<0.05, * p<0.1)

201

Table A.6 Marginal Effects of Multinomial Logistic Regressions on Vote for PP

DV: Vote for PP Variables Model 1 Model 2 Model 3 Model 4 Model 5 Model 6 Model 7

Age 0.003*** 0.002*** 0.002*** 0.002*** 0.002*** 0.002*** 0.002*** (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Gender -0.004 -0.006 -0.007 -0.006 -0.009 -0.006 -0.006 (0.006) (0.008) (0.008) (0.006) (0.008) (0.006) (0.006) Size Town/City -0.002 -0.005** -0.005** -0.002 -0.005** -0.002 -0.002 (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) (0.002) Education (Ref. Higher) Low -0.026** -0.024* -0.023 -0.015 -0.012 -0.015 -0.016 (0.012) (0.014) (0.014) (0.012) (0.014) (0.012) (0.012) Middle -0.008 -0.005 -0.006 -0.002 0.003 -0.002 -0.002 (0.010) (0.011) (0.011) (0.009) (0.011) (0.009) (0.009) Ideology 0.100*** 0.097*** 0.097*** 0.096*** 0.092*** 0.096*** 0.096*** (0.001) (0.002) (0.002) (0.001) (0.002) (0.001) (0.001) Unemployment -0.008 -0.004 0.002 (0.010) (0.010) (0.010) Income -0.003 -0.004* -0.003 (0.002) (0.002) (0.002) Interest Politics 0.014*** 0.013*** 0.014*** 0.014*** (0.003) (0.004) (0.003) (0.003) Abstention -0.079*** -0.070*** -0.079*** -0.079*** (0.010) (0.011) (0.010) (0.010) Economic Voting Sociotropic -0.041*** -0.051*** (0.015) (0.015) Pocketbook -0.041*** -0.027* (0.015) (0.015) Political Crisis -0.068*** -0.085*** -0.068*** -0.069*** (0.012) (0.013) (0.012) (0.012) Political Crisis X Ideology -0.156*** (0.029) Political Crisis X Age -0.075*** (0.013)

Controls yes yes yes yes yes yes yes Observations 9,251 7,335 7,350 9,202 7,280 9,202 9,202 Standard errors in parentheses (*** p<0.01, ** p<0.05, * p<0.1)

202

Figure A.1 Average Marginal Effects of Social Class on Probability to Vote for PP, PSOE, Podemos, and Ciudadanos

Notes: Results are based on Model 1 in Tables A.4, A.5, A.6 and A.7. The class scheme is constructed based on Oesch’s class divisions (see Oesch 2006). The results show that socio-cultural professionals, the social class usually associated with the New Left, are overrepresented in Podemos, while managers, who tend to vote liberal parties, are clearly overrepresented in Ciudadanos. Production workers, the class found to support the populist right in some countries remain overrepresented by the social-democratic party. Small business owners are the most overrepresented class in PP.

203

Appendix Ch. 6: Conflict Activation in Times of Crisis: The Case of Catalonia

Tables and Figures

Figure A.1 Predicted probabilities of support for independence over time and having voted for CiU vs. other parties in previous elections, 2008-2013

Note: Model results not shown, available upon request. Main dependent variable is support for an independent state. Controls include: age, gender, education, national sentiment, left-right scale, province and size of municipality.

204

Table A2: Logistic regression (odds ratio) of Deserters versus Loyals of CiU, 2010-2012.

Variables (1) (2) (3) (4)

Pro-Independence 11.054*** 3.674** 4.057*** 2.924** (4.761) (1.893) (2.070) (1.539)

Salience 1.756** 1.575 1.249 (0.459) (0.483) (0.448) National Sentiment (ref: Only/More Spanish) Dual 3.074 9.501 (4.774) (18.323) More Catalan 10.523 27.315 (17.373) (55.172) Only Catalan 10.820 23.138 (18.006) (47.494)

Evaluation A. Mas 1.122 1.111 (0.116) (0.120)

PID 3.423*** 3.705*** (1.236) (1.483)

Left-Right 2.206*** 2.410*** 2.316*** 2.521*** (0.426) (0.454) (0.525) (0.531)

Age 1.033*** 1.037*** 1.033** 1.032** (0.013) (0.013) (0.014) (0.014)

Constant 0.003*** 0.000*** 0.000*** 0.000*** (0.005) (0.000) (0.000) (0.000)

Controls ✔ ✔ ✔ ✔

Observations 222 221 222 221 Notes: Standard errors in parentheses (*** p<0.01, ** p<0.05, * p<0.1). Other controls included (not shown): gender, education, size of the municipality, province. Source: ICPS, survey 2012.

205

Figure A2: Share of ‘federalists’ that would vote ‘yes’ in a hypothetical referendum of independence, as % of federalists and total, 2011-2013

Source: CEO, 2006-2013.

206

Table A3: Regression results on support for independent state by political grievances, 2010-2013

Variable (1) (2) (3) (4)

Satisfaction w/ Democracy 0.068*** 0.030*** 0.029*** -0.001 (0.005) (0.004) (0.004) (0.011) Salience 0.805*** (0.0751) National Sentiment (ref: Only/More Spanish) Dual 0.014* -0.002 -0.006 (0.008) (0.010) (0.034) More Catalan 0.348*** 0.333*** 0.109** (0.011) (0.012) (0.044) Only Catalan 0.734*** 0.730*** 0.635*** (0.010) (0.012) (0.044)

Left-Right -0.013*** -0.012*** (0.003) (0.003)

Age -0.002*** -0.002*** (0.000) (0.000)

Nat. Sent. * Stf. Dem (ref: Only/More Spanish) Dual * Stf. Dem 0.001 (0.012) More Catalan * Stf. Dem 0.079*** (0.015) Only Catalan * Stf. Dem 0.033** (0.014)

Constant 0.051*** -0.083*** 0.152*** 0.237*** (0.018) (0.016) (0.038) (0.047)

Controls û û ü ü Region F.E. ü ü ü ü Wave F.E. ü ü ü ü

Observations 23,141 22,768 19,955 19,955 R-squared 0.064 0.425 0.444 0.446 Notes: Standard errors in parentheses (*** p<0.01, ** p<0.05, * p<0.1). Other controls included (not shown): gender, education, size of the municipality. Source: CEO, 2010-2013

207

Figure A3: Predicted Probabilities of voting for each party by national identity in different elections.

Notes: 2015 elections not included for ERC and CiU because they went in a pro-independence coalition together (JxSI). ICV includes CSQP in 2012 and CeC in 2015. Missing categories (e.g. ‘only catalan’ for C’s) are due to low number of observations. Controls include: Left-Right scale, gender, age, education, province and size of municipality. Regression results available upon request. Source: CIS post-electoral studies (2660; 2857; 2970; 3113; 3202).

208

Figure A4: Average scores for central government evaluation over time in Catalonia

Note: Scale from 0 to 10. Questions: “Evaluate the central government” (ICPS); “Trust in the Central Government” (CEO). Sources: ICPS, 1991-2016; and CEO, 2009-2013.

209

Figure A5: Marginal effects of being an ‘activated’ vs an ‘active’ voter by having voted for a pro- secessionist, 2006-2017

Notes: Active voters are those individuals who claim to vote in every election, whereas activated voters are those who claim to never or rarely do so. Pro-independence parties include CiU, ERC and CUP (in 2015 coalition JxSI and in 2017 CiU becomes JxCat). Controls include: Left-Right scale, gender, age, education, national identity, province and size of municipality. Regression results (logistic) available upon request. Source: CIS post-electoral studies (2660; 2857; 2970; 3113; 3202).

210

Figure A6: Marginal effects of being an ‘activated’ vs an ‘active’ voter by national identity, 2006- 2017

Notes: Active voters are those individuals who claim to vote in every election, whereas activated voters are those who claim to never or rarely do so. Controls include: Left-Right scale, party voted in last elections, gender, age, education, province and size of municipality. Regression results (logistic) available upon request. Source: CIS post-electoral studies (2660; 2857; 2970; 3113; 3202).

211

Figure A7: Predicted probabilities of belonging to each category of national identity by education levels.

Notes: Education categories are recoded for simplification purposes. Professional training at medium levels (grado medio) is coded into Secondary II and higher levels (grado superior) in Tertiary. Controls include: Year fixed effects, gender, age, education, province and size of municipality. Regression results (multinomial logit) available upon request. Source: ICPS 1991, 2013.

212

213